Month: July 2018

Article originally posted on InfoQ. Visit InfoQ

The Dangers of Dogmatism

But isn’t Scrum all I need to do to be Agile? Maybe, but not likely as no single framework is the answer to any complex organization’s Lean Agile Transformation, especially when reaching beyond the borders of IT to encompass overall organizational business agility. 

Still, there are those dogmatists and framework salesmen that will claim that Scrum will solve all your team level problems or that their Scaling framework fits all organizations’ scaling needs. While gains will be realized, the truly transformative gains that launch your organization into the high-performance 200x better than their competition requires a pragmatic Agnostic Agile mindset-driven approach that is truly customized to your teams and organization as a whole.  

So why is that? Well first let’s look at how a dogmatic, one-size-fits all approach hobbles an organization’s Lean Agile Transformation. The key reason is that it doesn’t recognize or address your organization’s uniqueness. Even in the same industry, organizations have their own unique history, culture, challenges and goals. 

An organization does itself a disservice by implementing in its entirety a given framework just because they have been convinced by some coach or framework salesmen that this works everywhere else so it’s perfect for you be it the Spotify model, Scrum, SAFe, LeSS, or whatever framework they are pushing. 

Have they spent any time truly understanding your history, culture, challenges, and goals? Yahoo and Google are both in the search engine and information aggregation business, but both are very different organizations. Furthermore, even within the same organization, groups and departments may have very different needs. Maybe the widget department should use Kanban while the grommet department uses a Scrum/XP blend. 

Secondly, a dogmatic, one-size-fits all approach does not allow for a healthy, unfettered Lean Agile evolution and maturing at both the team and organizational level. For example, a classic Command and Control enterprise may initially only be ready for a highly structured Scaling framework like SAFe; however it may evolve to the point where a far less structured model, such as Scrum@Scale or the Spotify Model would be more appropriate. In this case the organization needs to be made aware of lighter weight approaches and aided in transitioning to the best approach for them.  

Additionally, a dogmatic, one-size-fits all approach squelches innovation, learning and runs contrary to the Agile Principle of “At regular intervals, the team reflects on how to become more effective, then tunes and adjusts its behavior accordingly”; for a dogmatic approach constrains the ability both at the team and organizational levels to “adjust its behavior accordingly”.  Sadly, a team’s natural evolution and maturation can be seen as a violation of the playbook or framework guide and be nipped on the bud by a dogmatic coach who forces the team to do Agile by whatever book is their ‘bible’. Insisting that this is a ‘SAFe Shop’ and ‘we only do what is in the playbook’ and that anything else is ‘Scrum-Butting’ squelches any evolution and does the teams and organization a profound disfavor. 

A dogmatic framework-centric approach also squelches team and organization independence by making the team and organization dependent on their framework coach to make sure they are doing the framework right. Any deviation is seen as a failure that has them running back to the framework coach for a framework refresher to get them back on track. This fuels a cargo cult mentality where the team or organization sees following the framework guide as being Agile with little or no understanding of the Lean Agile Values and Principles underpinning the framework. 

Also lost is the fundamental truth that Lean Agile is a mindset not a specific framework and is about creating an empowering culture above all. It is no accident that the first value listed in the Agile Manifesto is “Individuals and Interactions over Processes and Tools”.

Now that we have seen why a dogmatic, one-size-fits all approach is a Lean Agile transformation antipattern and is actually not Agile at all, we will look at what an Agnostic Agile approach will do for us. 

We will do this by looking at how key Agnostic Agile principles facilitate and accelerate both the organization’s transformation as well as its Lean Agile evolution. Note this article contains my own interpretation and commentary on these principles. The principles with their original wording can be viewed here.

Principle 1: To put my customer first, making them independent

Besides tailoring a multi-framework approach that best fits the customer, putting the customer first means making the team and organization independent of the coach by teaching them to fish.

Healthy team evolution means that they go through three phases, commonly referred to as Shu-Ha-Ri.  This evolution can also be applied to the organization as a whole. 

• In Shu, the team members learn the frameworks they will be using. This is when a prescriptive approach is appropriate and called-for. 
• In Ha, the team is comfortable with the frameworks and starts being Agile. They are internalizing the Lean Agile Values and Principles and use that awareness to ‘break the rules’ or deal with grey areas in a way that is true to the Lean Agile Values and Principles.
• In Ri, the team is truly independent and self-coaching to the point where the specifics of a given framework become increasingly less important as they apply the now fully internalized Lean Agile Values and Principles to tailor and blend the frameworks to create a Lean Agile ecosystem unique to their needs, goals, environment and culture. This is how Scrumban and the Spotify model came to be. The team or organization is no longer practicing Scrum or SAFe, but are practicing OurAgile.

While the Agnostic Agile Coach understands that one needs to teach the frameworks, they know that a mono-framework approach is not optimal and that the Lean Agile Values and Principles must be taught in conjunction with the frameworks. Furthermore, they will emphasize that the values and principles are far more important that the specifics of the frameworks in question. Personally, I always advise team members and managers that the best approach to any problem, the best behavior in any situation, or the best answer to any question is the one that is truest to the Lean Agile Values and Principles. 

Once the team is comfortable with the framework and understands the importance of the Lean Agile Values and Principles, the Agnostic Agile Coach will mentor and coach the team and organization along the path of internalizing the principles and values. This facilitates their evolution to the Ha phase. The coach will challenge the team on their rational for breaking the rules or approach to grey areas to ensure that they have applied the values and principles in their decision making process as opposed to getting sloppy. 

Eventually the Agnostic Agile Coach will know that the team has truly internalized the values and principles and their decisions and approaches are true to the values and principles. The coach has taught them to fish and can ride off into the sunset as the team is now self-coaching. A good coach fires themselves, a bad coach becomes a permanent fixture and creates an unhealthy dependency on them. 

Principle 2: To do my best, complementing theory with practical experience


Do your coaches have practical experience? Your trainers? Your Scrum Masters? While book knowledge of the various frameworks is necessary, it is not sufficient for those who train, facilitate, mentor or coach. Practical experience tells you what really works, where the landmines are, what helps, what hinders, and where the rubber meets the road. Together the blend of theory and practice reduces the risk of doing harm and increases the ability to understand the customer’s unique aspects in order to craft an optimal approach for that customer. 

As an organization you can only benefit by your coaches, Product Owners and Scrum Masters having both a solid grounding in Lean Agile best practices as well as practical Agile experience. They will know what works and when to do it by the book, when not to and how to respond to situations where the framework used has no ready answers.

It is a concern that a team can be guided by someone with no experience and two days of training (CSM, PSM I); it is interesting that organizations that would never have placed an inexperienced person in the role of a waterfall PM, will place an inexperienced person in the role of Scrum Master. Even more of a concern is that someone with four days training (SPC4) could be leading an enterprise-wide scaled Agile transformation. In many cases the trainer who certified the Scrum Master or SPC4 isn’t required to have any practical Lean Agile experience either. This is tantamount to teaching someone how to drive without ever having driven.  Only if the trainer has practiced Lean Agile in the real world can they impart the knowledge required to learn practical, in the trenches Lean Agile techniques as opposed to dogma and Lego games. 

Principle 3: To tailor agility to context

As discussed in The Dangers of Dogmatism section, an organization is unique, so it both deserves and requires a Lean Agile approach that addresses and respects the organization’s uniqueness. This does not mean the Agnostic Agile Practitioner tolerates a lack of Agile growth, or faux Agile, however, the practitioner realizes that organizations cannot progress at the same rate and must be guided in a way that does not traumatize the organization, teams, and individuals by forcing too rapid a change. The practitioner’s experience and training lets them know when to push and when not to. That experience also allows the practitioner to tailor a blend of Lean Agile techniques that will work in the context of the organization’s reality.

Principle 4: To understand hindering constraints and work to remove them

Understanding the unique context and culture of the organization is key here, as is respecting where the organization, team or individual is.  Respect is important as the organization, team and individuals are getting through their day as is and have learned to survive and even thrive in their current environment and culture where wastes and other dysfunctions have become normalized and internalized. Brain science teaches us that small changes are far more likely to succeed. 

The Agnostic Agile Practitioner understands all of this and crafts their approach to removing constraints appropriately knowing when to tackle them head-on and when to flow around them like a stream around a rock. They can answer the ‘What’s in it for Them?’ question and ask them to ‘Help Me Help You’. Their pragmatic, agnostic mindset helps them avoid the dogmatic prescriptive ‘ram it down their throats’ approach which will hobble any transformation. 

Principle 5: To share, learn and improve

This is key to the growth of the individual, team, and organization. An Agnostic Agile Practitioner will work to create an environment, mindset and supporting processes that enable sharing, learning, and improving. Team and multi-team retrospectives, formal and informal training, guilds, hack days, show and tells, and any other means of learning and sharing those learnings will be pursued. This creation and sharing of knowledge fuels and facilitates an organization’s Lean Agile transformation.

An Agnostic Agile Practitioner will not feel threatened by another coach or practitioner who is more experienced or skilled, rather they will enlist that person as a mentor. No matter how skilled and experience, an Agnostic Agile Practitioner knows there is something that can be learned from any person they encounter no matter how ‘junior’ that person is. By learning and growing their knowledge, they increase their value to their organization or customer.

An Agnostic Agile Practitioner never presents themselves as the guru, for doing so creates a barrier to the sharing of knowledge and the growth of not only the team, team members, but the practitioner themselves. 

Principle 6: To respect frameworks and their practitioners

Agnostic Agile respects all frameworks, disciplines and their practitioners, as it is those frameworks that provide us with the tools we need to facilitate our team’s and organization’s transformation, as well as creating an empowering culture. 

The caveat is that the mono-framework practitioner must have an open, pragmatic attitude toward other frameworks and a willingness to work with practitioners of other frameworks forming multidisciplinary teams with the sole goal of doing what is best for the client and that respects where the client is in their organizational evolution. The Agnostic Agile practitioner will dialogue with the mono-framework practitioner in order to help them be more open to working in a multi-framework environment. By moving everyone to a more pragmatic, mutually respectful attitude, the organization benefits by having everyone focused on its transformation as opposed to being engaged in framework wars.

Principle 7: To acknowledge unknowns and seek help

An Agnostic Agile Practitioner is humble enough to know what they don’t know and when to go to other coaches and practitioners for help and advice as well as when to step aside and let another more experienced coach or practitioners step in if it is in the customer’s best interest. The Agnostic Agile Practitioner also knows when a framework specialist needs to be brought in, just as a Family Doctor knows when to refer a patient to a specialist. 

Principle 8: To never mislead and to never misrepresent

In all ways, an Agnostic Agile Practitioner acts with honesty and integrity in the best interest of the customer, team and individuals they interact with. If they honestly overestimated their ability, overlooked an option, or made an honest mistake, they will own it, admit to it and make amends.  Part of acting with integrity is not going along with a customer’s desire to pretend to be Agile in order to meet some outside requirement or executive directive to be an Agile shop when they personally have no intention of actually being Agile. Honesty and integrity are all-important in order to build the trust and gain the buy-in necessary for a successful transformation.

Principle 9: To remember that agility is not the end goal

It is a means to an end and there may be other means to the same end that make more sense for the customer’s growth and overall heath. The Agnostic Agile Practitioner assesses the organization’s culture, goals and issues to determine if other approaches or disciplines are more appropriate. For example, in fear-based or ultraconservative organizational cultures a Lean Agile approach is doomed to fail without first addressing the underlying psychological and sociological issues. In these cases, the organization is better served by bringing in organizational change and organizational behavior specialists prior to embarking on its Lean Agile transformation.

At the end of the day, the end goal should always be what is best for the customer, even if that means handing them over to other specialists and walking away.

Principle 10: To acknowledge that dogmatism is non-agile

I have already covered why dogmatism is not Agile and a violation of both the letter and spirit of the Agile Manifesto in the first section of this article. The dangers of dogmatism to the organization have also been covered.

The one thing to add is to warn against being so dogmatic about being pragmatic and agnostic that you view and treat mono-framework dogmatists with distain. An Agnostic Agile Practitioner accepts where people are and that they are sincere in their belief that they are being Agile. So it is necessary to treat them with the respect that is due to any person and engage in constructive and respectful dialogue using the Agile Manifesto as the common ground.  

Principle 11: To recognize that there is more to agile than agile

This one is related to principle 9. There are many fields an Agnostic Agile Practitioner can and should draw upon if the situation warrants it. These include Lean, Design Thinking, Management 2.0 and 3.0, organizational behavior, psychology, etc. A holistic, multimodal approach that best serves the customer is the order of the day.

Principle 12: To give to the community as it has given to me

An Agnostic Agile Practitioner acknowledges that the community has given them knowledge and support over the years. In their gratitude, they give back to the community by writing, posting, mentoring, and offering support as needed. They make time for their fellow practitioners and for those who wish to embark on their Agile journey. This benefits the customer as the practitioner grows their own knowledge in the process of giving back, plus the overall level of knowledge available in the Agile community grows.

Summary

An agnostic approach to Lean Agile will accelerate an organization’s Lean Agile transformation by crafting an approach that addresses that organization’s unique culture, goals, and issues.  Furthermore, by focusing on the Lean Agile values and principles as opposed to specific frameworks, an agnostic approach aids the organization, teams and individuals develop and internalize a Lean Agile mindset.  The combination of a pragmatic, customer-centered approach and the emphasis on ‘being Agile as opposed to doing Agile’ avoids the dangers of a prescriptive dogmatic framework-centric approach. 

For more on the Agnostic Agile approach, contact the author as well as visit the website.

About the Author

Tim Guay is well-versed in Lean Agile best practices with a proven track record as a practitioner, trainer and coach. He is experienced in introducing Agile and Lean into organizations, providing Agile and Lean training and coaching, as well as introducing DevOps best practices. He has been a practitioner, Agile trainer and coach to a wide-range of organizations from start-ups to Fortune 50 corporations since 2002. Tim has presented at conferences, developed courses and published articles on a wide-range of Lean Agile topics. He sits on ICAgile’s DevOps Track Certification Committee as was an early signatory of the Agnostic Agile Manifesto.  

You may contact Tim at timothyguay@yahoo.com and feel free to connect with him on LinkedIn where you can access his many Pulse articles and conference slidedecks.

Posted on mongodb google news. Visit mongodb google news

The most important aim of this analysis is to define, describe, and forecast the global Cloud-based Database market based on type, end-use industry, as well as developing regions. The global Cloud-based Database market analyzing the most potent regions around the globe Oceanian Sub-Region, North America, Latin America, The Middle East, Europe, Asia-Pacific and Africa. The global Cloud-based Database analysis includes detailed information regarding the significant elements like drivers, opportunities, restraints, and challenges influencing the increase of the ion exchange resins market over the world. The Cloud-based Database report strategically analyzes market segments related to human growth trends, growth prospects, and donation to the overall market.

The analysis aids an individual by providing useful info regarding product offerings, industry overviews, and sales segmentation of the greatest global Cloud-based Database market players. Additionally, it forecasts the growth of the vital market players throughout global Cloud-based Database market SWOT analysis and different analyzing techniques. The worldwide Cloud-based Database market report examines the growth of the essential market players with the assistance in these recent expansions on the industry. Further, this report is directed by Cloud-based Database Application/end users, and types predicated on historical and projected trade share and compounded annual growth rate (CAGR in percent) with size (Units) and Revenue (Mn/Bn USD).

Get Sample of this Report at: http://emarketresearch.us/global-cloud-based-database-market/#Request-Sample

Leading Players: SAP, Couchbase, Rackspace Hosting, Cassandra, Oracle, Amazon Web Services, Microsoft, Google, Salesforce, Tencent, Alibaba, IBM, MongoDB and Teradata

World Wide Cloud-based Database market (Countless Units) and Revenue (Mn/Bn USD) Economy Divide by Product Type:

SQL Database
NoSQL Database

Report coordinates by Applications with projected and historical Cloud-based Database market share:

Small and Medium Business
Large Enterprises

Geographically this record is split into several important regions, together with production, consumption, and increased pace of Cloud-based Database in those regions, by 2012 to 2023 (forecast), covering Oceanian Sub-Region, North America, Latin America, The Middle East, Europe, Asia-Pacific and Africa share (percent) and also CAGR for its projections 2018 to 2023.

Goals of the research:

➜ Describe the latest Cloud-based Database improvements, market stocks, and approaches utilized by most significant players;

➜ Study on the regions which can be anticipated to see the fastest growing growth in the forecast interval;

➜ To reevaluate chances for stakeholders by identifying high-growth sections of the Cloud-based Database market;

➜ To ascertain and predict the client participation solutions market, Cloud-based Database business dimensions, verticals, and regions from 2018 to 2023, and also analyze different macro- and microeconomic variables which influence market development;

➜ To attain research-based business decision and put on weight to present and marketing stuff and gain a competitive understanding of Cloud-based Database market-leading players;

➜ The Cloud-based Database report shows the way the rigorous emission management standards will induce the international market;

➜ To profile essential market players and supply comparative evaluation based on company overviews, merchandise offerings, regional existence, company plans, to Comprehend the competitive landscape;

➜ To provide comprehensive advice about important Factors like Cloud-based Database market drivers, restraints, opportunities, and challenges influencing the development of the market;

➜ Study on the type that’s expected to control the Cloud-based Database ;

➜ Analyzing a variety of viewpoints of this market with the assistance of all both porter’s five forces analysis;

➜ To monitor and analyze aggressive progress including Cloud-based Database mergers & acquisitions, contracts & agreements, joint ventures, partnerships, and strategic positions from the client participation solutions market;

For Any Query, Speak to Our Expert: http://emarketresearch.us/global-cloud-based-database-market/#Inquiry-Before-Buying

The Cloud-based Database market report suits various stakeholders within this business, for example, shareholders, manufacturers, vendors and providers for equipment, government companies, consulting and research firms, new entrants, and fiscal analysts. Various plan matrices utilized in assessing the global Cloud-based Database market would supply stakeholders important inputs to create strategic decisions so.

Afterward, the Cloud-based Database report targets worldwide leading top industry players together with advice like company profiles, product specifications and pictures, sales, market share and contact info. Whats more, the more global Cloud-based Database industry growth trends and stations have been examined.

Article originally posted on mongodb google news. Visit mongodb google news

Article originally posted on Data Science Central. Visit Data Science Central

With Beginning Data Science with Python and Jupyter, get to grips with the skills you need for entry-level Data Science. You’ll learn about some of the most commonly used libraries that are part of the Anaconda distribution, and then explore machine learning models with real datasets to give you the skills and exposure you need for the real world. This book will finish up by showing you how easy it can be to scrape and gather your own data from the open web, so that you can apply your new skills in an actionable context.

Getting started with data science doesn’t have to be an uphill battle. This step-by-step guide is ideal for beginners who know a little Python and are looking for a quick, fast-paced introduction. This book covers every aspect of the standard data-workflow process within a day, along with theory, practical hands-on coding, and relatable illustrations.

Who this book is for:

This book is ideal for professionals with a variety of job descriptions across large range of industries, given the rising popularity and accessibility of data science. You’ll need some prior experience with Python, with any prior work with libraries like Pandas, Matplotlib and Pandas providing you a useful head start.

Key Features:

• Get up and running with the Jupyter ecosystem and some example datasets
• Learn about key machine learning concepts like SVM, KNN classifiers and Random Forests
• Discover how you can use web scraping to gather and parse your own bespoke datasets

What you will learn:

• Get up and running with the Jupyter ecosystem and some example datasets
• Learn about key machine learning concepts like SVM, KNN classifiers, and Random Forests
• Plan a machine learning classification strategy and train classification, models
• Use validation curves and dimensionality reduction to tune and enhance your models
• Discover how you can use web scraping to gather and parse your own bespoke datasets
• Scrape tabular data from web pages and transform them into Pandas DataFrames
• Create interactive, web-friendly visualizations to clearly communicate your findings

Get a preview of the book featuring Jupyter fundamentals.

Posted on mongodb google news. Visit mongodb google news

Article originally posted on mongodb google news. Visit mongodb google news

Article originally posted on InfoQ. Visit InfoQ

Bio

Joseph Williams is a former executive with Microsoft and Sun Microsystems, and he served as the Dean of the School of Business, Government and Economics at Seattle Pacific University before joining Commerce. He has authored three books and dozens of articles related to the ICT industry while he was a professor in the College of Business at Colorado State University in the 1990s.

Global Big Data Conference’s vendor agnostic Global Blockchain Conference is held on April 2nd, April 3rd, & April 4th 2018 on all industry verticals(Finance, Retail/E-Commerce/M-Commerce, Healthcare/Pharma/BioTech, Energy, Education, Insurance, Manufacturing, Telco, Auto, Hi-Tech, Media, Agriculture, Chemical, Government, Transportation etc.. ). It will be the largest vendor agnostic conference in Blockchain space. The Conference allows practitioners to discuss Blockchain through effective use of various techniques.

Posted on mongodb google news. Visit mongodb google news

Shares of Mongodb Inc (NASDAQ:MDB) have earned an average recommendation of “Buy” from the fifteen analysts that are currently covering the stock, MarketBeat reports. One investment analyst has rated the stock with a sell rating, five have issued a hold rating and nine have assigned a buy rating to the company. The average 12-month price target among brokers that have issued ratings on the stock in the last year is $50.20.

Several analysts have recently weighed in on the stock. Zacks Investment Research downgraded shares of Mongodb from a “hold” rating to a “sell” rating in a research report on Sunday, July 15th. Rosenblatt Securities started coverage on shares of Mongodb in a research report on Monday, July 9th. They issued a “buy” rating for the company. Citigroup started coverage on shares of Mongodb in a research report on Thursday, June 28th. They issued a “neutral” rating and a $56.00 price target for the company. Needham & Company LLC upped their price target on shares of Mongodb from $52.00 to $57.00 and gave the company a “buy” rating in a research report on Tuesday, June 12th. Finally, Stifel Nicolaus upped their price target on shares of Mongodb from $47.00 to $53.00 and gave the company a “buy” rating in a research report on Thursday, June 7th.

In other Mongodb news, insider Eliot Horowitz sold 9,262 shares of the business’s stock in a transaction on Monday, July 2nd. The shares were sold at an average price of $49.06, for a total value of $454,393.72. The sale was disclosed in a legal filing with the SEC, which is available through this link. Also, major shareholder Whale Rock Capital Management acquired 1,214,850 shares of the company’s stock in a transaction on Tuesday, May 8th. The stock was purchased at an average cost of $40.23 per share, for a total transaction of $48,873,415.50. The disclosure for this purchase can be found here. In the last quarter, insiders sold 156,789 shares of company stock valued at $8,952,336. 40.04% of the stock is owned by insiders.

A number of institutional investors and hedge funds have recently made changes to their positions in MDB. Bank of New York Mellon Corp acquired a new position in Mongodb in the 4th quarter valued at $548,000. BlackRock Inc. acquired a new position in shares of Mongodb during the 4th quarter worth $10,042,000. UBS Asset Management Americas Inc. acquired a new position in shares of Mongodb during the 4th quarter worth $879,000. Geode Capital Management LLC acquired a new position in shares of Mongodb during the 4th quarter worth $2,864,000. Finally, Goldman Sachs Group Inc. acquired a new position in shares of Mongodb during the 4th quarter worth $4,493,000. 26.89% of the stock is currently owned by institutional investors.

Shares of NASDAQ:MDB traded up $0.76 during trading on Friday, hitting $61.32. 18,979 shares of the company were exchanged, compared to its average volume of 821,274. The company has a market capitalization of $2.98 billion and a PE ratio of -35.45. Mongodb has a 12-month low of $24.62 and a 12-month high of $61.00.

Mongodb (NASDAQ:MDB) last issued its quarterly earnings data on Wednesday, June 6th. The company reported ($0.43) earnings per share for the quarter, hitting the consensus estimate of ($0.43). The company had revenue of $48.20 million for the quarter, compared to analysts’ expectations of $46.46 million. The business’s revenue was up 48.8% on a year-over-year basis. equities research analysts expect that Mongodb will post -2.22 EPS for the current fiscal year.

About Mongodb

MongoDB, Inc operates as a general purpose database platform worldwide. The company offers MongoDB Enterprise Advanced, a subscription package for enterprise customers to run in the cloud, on-premise, or in a hybrid environment; MongoDB Atlas, a cloud-hosted database-as-a-service solution; and Community Server, a free-to-download version of its database, which includes the functionality that developers need to get started with MongoDB.

Featured Story: Book Value Per Share in Stock Trading

Receive News & Ratings for Mongodb Daily – Enter your email address below to receive a concise daily summary of the latest news and analysts’ ratings for Mongodb and related companies with MarketBeat.com’s FREE daily email newsletter.

Article originally posted on mongodb google news. Visit mongodb google news

Article originally posted on Data Science Central. Visit Data Science Central

Guest blog by Yoel Zeldes.

This post describes:

• what the Gumbel distribution is
• how it is used for sampling from a discrete distribution
• how the weights that affect the distribution’s parameters can be trained
• how to use all of that in a toy example (with code)

In this post you will learn what the Gumbel-softmax trick is. Using this trick, you can sample from a discrete distribution and let the gradients propagate to the weights that affect the distribution’s parameters. This trick opens doors to many interesting applications. For start, you can find an example of text generation in the paper GANS for Sequences of Discrete Elements with the Gumbel-softmax Distribution.

Gumbel distribution

Introduction

Training deep neural networks usually boils down to defining your model’s architecture and a loss function, and watching the gradients propagate.

However, sometimes it’s not that simple: some architectures incorporate a random component. The forward pass is no longer a deterministic function of the input and weights. The random component introduces stochasticity, by means of sampling from it.

When would that happen, you ask? Whenever we want to approximate an intractable sum or integral. Then, we can form a Monte Carlo estimate. A good example is the variational autoencoder. Basically, it’s an autoencoder on steroids: the encoder’s job is to learn a distribution over the latent space. The loss function contains an intractable expectation over that distribution, so we sample from it.

As with any architecture, the gradients need to propagate to the weights of the model. Some of the weights are responsible for transforming the input into the parameters of the distribution from which we sample. Here we face a problem: the gradients can’t propagate through random nodes! Hence, these weights won’t be updated.

One solution to the problem is the reparameterization trick: you substitute the sampled random variable with a deterministic parameterized transformation of a parameterless random variable.

If you don’t know this trick I highly encourage you to read about it. I’ll demonstrate it with the Gaussian case. For many types of continuous distributions you can do the reparameterization trick. But what do you do if you need the distribution to be over a discrete set of values?

Read the full article with source code, here. 

DSC Resources

• Free Book: Applied Stochastic Processes
• Comprehensive Repository of Data Science and ML Resources
• Advanced Machine Learning with Basic Excel
• Difference between ML, Data Science, AI, Deep Learning, and Statistics
• Selected Business Analytics, Data Science and ML articles
• Hire a Data Scientist | Search DSC | Classifieds | Find a Job
• Post a Blog | Forum Questions

Posted on mongodb google news. Visit mongodb google news

Investment company Schf (gpe), Llc buys MongoDB Inc, sells PDC Energy Inc, Despegar.com Corp, Bellatrix Exploration during the 3-months ended 2018-06-30, according to the most recent filings of the investment company, Schf (gpe), Llc. As of 2018-06-30, Schf (gpe), Llc owns 1 stocks with a total value of $40 million. These are the details of the buys and sells.

For the details of SCHF (GPE), LLC’s stock buys and sells, go to http://www.gurufocus.com/StockBuy.php?GuruName=SCHF+%28GPE%29%2C+LLC

These are the top 5 holdings of SCHF (GPE), LLC

1. MongoDB Inc (MDB) – 798,088 shares, 100% of the total portfolio. New Position
2. Bellatrix Exploration Ltd (BXE) – 0 shares, 0% of the total portfolio. Shares reduced by 10000%
3. Despegar.com Corp (DESP) – 0 shares, 0% of the total portfolio. Shares reduced by 10000%
4. PDC Energy Inc (PDCE) – 0 shares, 0% of the total portfolio. Shares reduced by 10000%

New Purchase: MongoDB Inc (MDB)

Schf (gpe), Llc initiated holding in MongoDB Inc. The purchase prices were between $34.77 and $59.23, with an estimated average price of $44.83. The stock is now traded at around $61.43. The impact to a portfolio due to this purchase was 100%. The holding were 798,088 shares as of 2018-06-30.

Sold Out: PDC Energy Inc (PDCE)

Schf (gpe), Llc sold out a holding in PDC Energy Inc. The sale prices were between $45.53 and $65.32, with an estimated average price of $57.57.

Sold Out: Despegar.com Corp (DESP)

Schf (gpe), Llc sold out a holding in Despegar.com Corp. The sale prices were between $20.54 and $31.06, with an estimated average price of $25.79.

Sold Out: Bellatrix Exploration Ltd (BXE)

Schf (gpe), Llc sold out a holding in Bellatrix Exploration Ltd. The sale prices were between $0.96 and $1.68, with an estimated average price of $1.22.

Here is the complete portfolio of SCHF (GPE), LLC. Also check out:

1. SCHF (GPE), LLC’s Undervalued Stocks
2. SCHF (GPE), LLC’s Top Growth Companies, and
3. SCHF (GPE), LLC’s High Yield stocks
4. Stocks that SCHF (GPE), LLC keeps buying

Article originally posted on mongodb google news. Visit mongodb google news

Article originally posted on InfoQ. Visit InfoQ

You will be sent an email to validate the new email address. This pop-up will close itself in a few moments.

Article originally posted on InfoQ. Visit InfoQ

Jonathan Smart, head of working ways at Barclays, and Morag McCall, project management officer (PMO) at Barclays, spoke last month at the DevOps Enterprise Summit in London, about re-thinking the entire flow of work, from initial idea triage until releasing to production. This means introducing agility into how the application and services portfolio is managed, as well as changing the role of the PMO and finance departments.

Smart had spoken in previous years about how Agile and DevOps were successfully introduced at Barclays. Those transformations were focused on engineering but with the aim of helping reduce lead time to deliver new outcomes for the business. By now looking at the end to end flow of activities – starting from an initial business idea and ending with a new feature or story in production – it became clear there were much larger improvements to flow to be made on the activities before the engineering work actually started. The full value stream would start with monthly ideas triage, then quarterly steering, and annual approval before a new idea even could land a spot in the development backlog. Smart called this mismatch (between an up to 18 months long portfolio management process and the Agile engineering process the “urgency paradox”).

The revamped portfolio management process now starts with strategic objectives. This is the furthest, multi-year horizon business vision for the organization (the CEO’s top-5 objectives), or large programs like Brexit. These objectives are broken down into portfolio objectives with a 3 years view. A portfolio objective is a grouping of one or more portfolio epics. These yearly portfolio epics are divided into 3 months or shorter business outcomes (the everyday working unit) that are being constantly validated – with a limit of 4 business outcomes in progress simultaneously. Finally, features are delivered monthly and measured towards the progress of the corresponding business outcome (seen as experiments). According to Smart, it is now possible for an engineer at Barclays to trace the story they are working on any given day (multiple stories make up a monthly feature) all the way up to the CEO’s top-5 objectives.

Credit: Jonathan Smart and Morag McCall, Barclays

Crucially, these vertical initiatives map to horizontal (long lived) products and services. A single business outcome might require changes to multiple (long-lived) services and products. For example, a business outcome that needs changes to both the Fraud and Equity Trading services, which in turn might require changes to multiple products on each service. Nevertheless, the intent is to continuously reduce the number of dependencies, said Smart.

This new portfolio management approach does not mean the PMO role is outdated, much to the contrary. But portfolio management teams need to evolve towards a more agile approach to work planning and (self-)optimizing – possibly with the help from an Agile coach, and limiting work in progress to maximize the flow of business outcomes (“stop starting, start finishing”). Physical work boards can help visualize the work across the team, electronic boards should only be introduced later. Moving from a project mindset to (long-lived) product orientation (and corresponding long-lived product teams) is also critical. The end goal for portfolio management is to enable business value via an advisory role: consulting and supporting the rest of the organization, namely helping with dependency and release management (getting the work to flow, rather than planning work to be done).

According to Smart, finance also needs to change to enable a lean portfolio management with quarterly, rolling business outcomes. Funding value streams rather than projects, considering lightweight business cases, and making the procurement process more agile are some of the requirements.

Smart and McCall ended their talk with a set of recommendations to start improving portfolio management in large organizations:

• identify long-lived value streams and products
• identify quarterly business outcomes
• visualize and limit portfolio work in progress (WIP)
• focus on reducing lead time
• build agility by becoming an enabler

Video of the talk is available online from IT Revolution, the DevOps Enterprise Summit organizers.