Month: June 2018

It’s Not About the Chocolate

In his podcast, Akimbo, Seth Godin teaches us how to adopt a posture of possibility, change the culture, and choose to make a difference. Here are my takeaways from the episode.

  • A question that society often wrestles with is “What is culture and how do we define it?” Back in the 1930’s, economist John Maynard Keynes came up with a thesis that, for humans to be happy, they have to do everything they can to get more. With our abundance of productivity today, does that thesis still make sense?
  • The researches on many cultures tell us that culture can be defined as the common story we tell ourselves. If we tell ourselves a story that we view as normal, that is the culture. In culture, we do it because we think it makes us happy. We do it because we think it makes us fit in or stand out. Culture is what people do, people like us. If the story we tell ourselves is not making us happy, we need to tell yourself a different story.
  • There are two ideas at work here. First, we can change the story we tell ourselves. If we are open to finding a new story, one about participation or sufficiency or meaning and we can surround ourselves with ideas or even better people, we can build a new culture. That culture can make us happier, more productive, more engaged.
  • The second idea is about “having more.” Shawn Askinosie’s chocolate story is not about “having more.” Rather it is about achieving “enough” and use it as a lever and opportunity to create a new culture, a culture where it’s more likely all participants are going to be happy with this story.
  • Some might say the capitalism is enabling culture. Instead, we should think about using capitalism to enable a culture. A culture where the purpose of capitalism is to leverage trade, productivity, and engagement to make us happier.
  • We have a dilemma as people who want to influence the culture because a lot of changes that we would like to make cannot be learned or happen overnight. All the things that are important to us, that we pay for, that we wait in line for, that we remember fondly, are things that were hard-won. Those things involve nuance and sophistication. Often, they required a bit of a leap of faith, trial and error, and digging in deeper. If we are going to try to change the culture, we must make a choice; the choices are race-to-the-top or race-to-the-bottom. The problem of the race-to-the-bottom is you might win, or, worse, come in second and never mattered.
  • We do not need everyone. We do not even need a lot of people. We just need a few people who care, a few people who will enroll in the journey who need to be seen and are willing to see. If we can produce that work gradually, drip-by-drip and day-by-day, maybe we can get under people’s skin. If we can transform them, we can raise the standard. Fortunately, there are people in the world who care enough to be meaningful and specific and focused and difficult as opposed to simply chasing more.

Updated Machine Learning Templates for Python and R

As I work on practicing and solving machine learning (ML) problems, I find myself repeating a set of steps and activities repeatedly.

Thanks to Dr. Jason Brownlee’s suggestions on creating a machine learning template, I have pulled together a set of project templates that can be used to support regression ML problems using Python and R.

Version 3 of the templates contain several minor adjustments and corrections to address discrepancies in the prevision versions of the template.

You will find the templates in the Machine Learning Project Templates page.

You can also check out the sample HTML-formatted report here on GitHub.

Python Templates and Reports

R Templates and Reports

卓越

(從我的一個喜歡與尊敬的作家,賽斯 高汀

如果你知道,

你可以通過顧客的眼睛看到這個世界,

你真的很在意…

你會怎麼做?

這是一個創造卓越的簡單測試。

那麼,如果我在Orbitz的電話線上呆等了五十六分鐘,那公司的總裁是否會知道?這是否是我們想看到的結果?

那位設計與運送投票機的工程師是否知道那機器已是可破解的?

那位水電工完成了這項工作,並將冷/熱控制置於相反的位置…他是否也足夠的在乎?

卓越可以削減了官僚主義,現狀,和藉口,並提出一個簡單的問題:

如果你知道會這樣,你會怎麼做?

Structure Plan, Part 2

In the book, Bare Bones Change Management: What you shouldn’t not do, Bob Lewis explained the seven must-have elements for any change management effort to have a chance of succeeding. Here are my takeaways from one of the topics discussed in the book.

[https://smile.amazon.com/Bare-Bones-Change-Management-shouldnt-ebook/dp/B004UOH5XW/]

A structure plan describes how the organization is put together to support the change. The plan includes:

How to organize: To support a change, everyone in the organization needs to know what his/her part of the organization needs to do. If an organization’s basic structure is not consistent with the change we are hoping to install, it will prevent the change from taking place. This component of the plan addresses two key elements: a clear organizing strategy and the specific reporting relationships.

There needs to be clear organizing strategy. Some examples of organizing could be by product, customer segment, size, demographics, sector, channel, or function. After we make the basic organizing decision, we need to define the reporting relationships. The relationships decision can have two folds: 1) flat or deep, and 2) realign, reorganize, or integrate.

Some firms celebrate victory right after the reorganization, but that is too early. A reorganization is not the same as making the change happen.

Facilities: With the organization changes taking place, we will need to plan for the movement and positioning of the people. This component lay out the physical nature of the workplace – who sits near whom and which departments and workgroups are in proximity.

Governance: Every organization has a culture, and the culture defines how we make decisions. This plan component describes both the official process and the informal ones that precede it and surround it.

Accounting: The accounting system tracks the ownership of which expenses and revenue attributed to the change. Bob recommends doing everything we can to keep all political considerations out of the accounting changes.

Compensation: This plan component puts considerations around two concerns: whose compensation has to change, and how evaluation criteria need to change. The behaviors, attitudes, skills, and intangibles the company values for the change must be expressed through training and awareness campaigns.

Multi-Class Classification Model for Faulty Steel Plates Using R

Template Credit: Adapted from a template made available by Dr. Jason Brownlee of Machine Learning Mastery.

Dataset Used: Faulty Steel Plates

Dataset ML Model: Multi-Class classification with numerical attributes

Dataset Reference: http://archive.ics.uci.edu/ml/datasets/steel+plates+faults

One potential source of performance benchmarks: https://www.kaggle.com/uciml/faulty-steel-plates

INTRODUCTION: This dataset comes from research by Semeion, Research Center of Sciences of Communication. The original aim of the research was to correctly classify the type of surface defects in stainless steel plates, with six types of possible defects (plus “other”). The Input vector was made up of 27 indicators that approximately the geometric shape of the defect and its outline. According to the research paper, Semeion was commissioned by the Centro Sviluppo Materiali (Italy) for this task and therefore it is not possible to provide details on the nature of the 27 indicators used as Input vectors or the types of the 6 classes of defects.

CONCLUSION: The baseline performance of the seven algorithms achieved an average accuracy of 69.69%. Three algorithms (Bagged CART, Random Forest, and Stochastic Gradient Boosting) achieved the top three accuracy scores after the first round of modeling. After a series of tuning trials, Stochastic Gradient Boosting turned in the top result using the training data. It achieved an average accuracy of 77.78%. Using the optimized tuning parameter available, the Stochastic Gradient Boosting algorithm processed the validation dataset with an accuracy of 77.20%, which was slightly below the accuracy of the training data. For this project, the Stochastic Gradient Boosting ensemble algorithm yielded consistently top-notch training and validation results, which warrant the additional processing required by the algorithm.

The HTML formatted report can be found here on GitHub.