Automated Machine Learning

This week I was invited to give a talk at the Haifa Tech Talks meetup. In this talk I presented an introduction to Automated Machine Learning (Auto ML).
I discussed some approaches to face the challenges of Auto ML, including Data Preprocessing, Hyperparameter Tuning and Algorithm Selection.
I also presented two popular tools for Auto ML: Auto-sklearn (based on Bayesian Optimization) and TPOT (based on Genetic Algorithms).

About the speaker:
Hayim Makabee has 25 years experience with Software Development, and in the last 10 years has specialized in the field of Data Science.
He is currently the CEO at KashKlik, an advanced Influencer Marketing platform.
Hayim provides consultancy and mentorship to startup companies in the fields of Machine Learning and Predictive Analytics.

You can see my slides and video below.

Here is the video of the talk (in Hebrew):

This talk was hosted by the Haifa Tech Talks meetup group. Special thanks to Roman Levin who organized the event.

Posted in Data Science, Machine Learning, Research | Tagged , , | 1 Comment

Managing your Reputation

This week I was invited to give talk to a group of Brazilian Olim (new immigrants) in Israel. In this talk I shared useful guidelines about how to manage and develop your personal reputation. I focused on providing practical advice about how to create opportunities by generating value to the people in your professional network.

You can see my slides and video below.

Here is the video of the talk (in Portuguese):

Special thanks to Carol Hauser Slapak who organized the event and invited me to give this talk.

Posted in Efficacy, Social Networks | Tagged , | 1 Comment

Coronavirus, Black Swans and Worst-Case Scenarios

blackswanThe concept of Black Swans was popularized by the author Nassim Taleb. According to his definition:

“What we call here a Black Swan (and capitalize it) is an event with the following three attributes.

First, it is an outlier, as it lies outside the realm of regular expectations, because nothing in the past can convincingly point to its possibility. Second, it carries an extreme ‘impact’. Third, in spite of its outlier status, human nature makes us concoct explanations for its occurrence after the fact, making it explainable and predictable.

I stop and summarize the triplet: rarity, extreme ‘impact’, and retrospective (though not prospective) predictability. A small number of Black Swans explains almost everything in our world, from the success of ideas and religions, to the dynamics of historical events, to elements of our own personal lives.”

It is clear that the current spread of the Coronavirus is an example of a Black Swan. It was completely unexpected, it has extreme consequences and is now being explained. In practice what it means is that our reality became much worse than any worst-case scenario.

Worst-Case Scenarios

When people make plans, a traditional way to develop a solid strategy is to consider several potential scenarios. Because we are dealing with the future, we cannot be absolutely sure about anything, so creating possible scenarios sounds like a good way to deal with the uncertainties. One of these scenarios is the worst-case scenario:

“A worst-case scenario is a concept in risk management wherein the planner, in planning for potential disasters, considers the most severe possible outcome that can reasonably be projected to occur in a given situation. Conceiving of worst-case scenarios is a common form of strategic planning, specifically scenario planning, to prepare for and minimize contingencies that could result in accidents, quality problems, or other issues.”

However, when there is a Black Swan, like the current Coronavirus pandemic, reality may become worse than the worst-case scenario. The Black Swan is by definition an event totally unexpected, and as such it could not have been part of our plans.

Goals, Plans and Attributes

The purpose of a plan is to reach some goal. But when there is a Black Swan, all the plans become irrelevant, because the goals themselves need to be changed. For example, instead of having the goal to grow their revenue or increase their sales, at this moment many companies are just trying to survive.

In order to face unexpected situations, we do not need plans, but we do need relevant attributes. In the case of software systems these are normally called “non-functional attributes”, because they are not directly related to the function performed by the system. These non-functional attributes are properties that all systems should ideally have. Examples are Robustness, Resilience and Fault-Tolerance.

Thus we do not need a predefined plan to handle the Coronavirus crisis, and it should not be taken in consideration as a potential worst-case scenario. What we do need is to develop the attribute of Robustness in order to survive this or any other Black Swan that we will never be able to predict.

Posted in Antifragility, Efficacy | Tagged , | Leave a comment

Welcome to Chaos: Corona, Bat Soup and the Butterfly Effect

butterflyThe natural tendency of humans is to look for order. In general, the progress and development of technology and society have enabled us to live in a world that is much more organized and predictable than the one of our ancestors. However there are moments in which this order is disrupted and things become chaotic. We are living one of such moments now, with the Corona virus spreading fast across countries.

But what looks like chaos is not necessarily the total lack of order. Actually, chaotic states may also be studied, and this is the purpose of Chaos theory. According to Wikipedia:

“Chaos theory is a branch of Mathematics focusing on the study of chaos—states of dynamical systems whose apparently-random states of disorder and irregularities are often governed by deterministic laws that are highly sensitive to initial conditions. Chaos theory is an interdisciplinary theory stating that, within the apparent randomness of chaotic complex systems, there are underlying patterns, interconnectedness, constant feedback loops, repetition, self-similarity, fractals, and self-organization.”

One of the most popular concepts of Chaos theory is the Butterfly Effect:

“The Butterfly Effect, an underlying principle of chaos, describes how a small change in one state of a deterministic nonlinear system can result in large differences in a later state (meaning that there is sensitive dependence on initial conditions). A metaphor for this behavior is that a butterfly flapping its wings in China can cause a hurricane in Texas.”

Corona Virus

bat soupAt this moment, the Corona virus is disrupting the life of hundreds of millions of people globally. Factories have been closed, travelers have cancelled their trips, conferences have been postponed, children are studying from home.

There is a rumor that everything started with a Chinese guy eating bat soup. Imagine that: one person eating bat soup in China causes Billions of dollars of losses in the economy worldwide. Exactly like the Butterfly Effect.

But it is not really important if the Corona virus started with a bat soup. The fact is that at some point a single person got infected accidentally, and now all the world is suffering the consequences. This is Chaos theory, in practice.

1 Comment

Time immigrants: facing the unknown

time_travelAn immigrant is a person that makes the decision to leave his native country and move to a foreign country. The immigrant is conscious that he will face many difficulties and will have to adapt to a new reality. He may need to learn a new language, and probably will have to acquire new skills. When a person immigrates, he decides to face the unknown. He leaves a world that is very familiar and well understood, and moves into a new environment that is full of challenges.

Today we are living in a world that is changing extremely fast. New sophisticated technologies are constantly replacing other technologies that become obsolete. Some traditional professions are disappearing while new specializations are being created. College students are not sure if they will be able to find good jobs when they graduate, and at the same time many companies have difficulties hiring qualified professionals with the right expertise.

We are not able to predict how exactly our lives will change in the next decades, but we know for sure that the future will be very different from the present. We will probably face many difficulties and have to adapt to a new reality. We may need to acquire new skills. We are leaving a world that is very familiar and well understood, and moving into a new environment that is full of challenges. We are not immigrating geographically, but we are time immigrants.

Are you ready to time-immigrate? What is your strategy to succeed in a world that will be drastically different from the one you know?

Posted in Efficacy | Tagged | Leave a comment

On Brainstorms and Criticism

This is the definition of Brainstorming from the Wikipedia:

“Brainstorming is a group creativity technique by which efforts are made to find a conclusion for a specific problem by gathering a list of ideas spontaneously contributed by its members.

In other words, brainstorming is a situation where a group of people meet to generate new ideas and solutions around a specific domain of interest by removing inhibitions. People are able to think more freely and they suggest as many spontaneous new ideas as possible. All the ideas are noted down without criticism and after the brainstorming session the ideas are evaluated.”

In the quote above, the emphasis on “without criticism” is mine. As we know, it is not a natural tendency for people to avoid making criticism. They may even be able to refrain from making an open criticism, but in general it is quite hard for people to identify with the innovative ideas being proposed by others.

Therefore, even if during the brainstorming session the ideas are not criticized, most of them will be immediately discarded when “after the brainstorming session the ideas are evaluated.”

This is very well depicted in the cartoon below:


Posted in Efficacy | Tagged | 2 Comments

Learning From Others’ Mistakes

124191160-new-year-from-2019-to-2020-colored-3d-numbers-with-shadow-on-a-white-backgroundToday is the last day of the year 2019. At this very moment, most people are thinking about how they are going to change next year. They are making plans for 2020. But I’m making a retrospective: I’m thinking about how did I change in 2019.

I believe that the most important change in my attitude has been to focus on learning from others’ mistakes. Previously, I was very excited about reading success stories. I discuss this change in my approach in my previous post: “Should we Learn from Success Stories?

I also understood that avoiding mistakes is what really increases our probability of success. I stopped trying to follow rules or to adopt a formula to be successful. I discuss this idea in my previous post: “On Rules and Probabilities

Following this change in my attitude, I started asking the question: What is the science behind bad decisions? How can we really avoid making mistakes?

Learning from Chess Games

One of the most interesting articles I read on this topic was “Data Mining Reveals the Crucial Factors That Determine When People Make Blunders“, from MIT Technology Review.

This article is based on a scientific investigation using a database of 200 million chess games. The researchers have analyzed the games to detect whether a player has made a mistake, and what were the factors that have caused this mistake.

These are some of the main insights from this article:

“Decision making is influenced by the complexity of the situation, the skill of the decision maker, and the time pressure.”

“The amount of time spent on a decision is a factor in blundering, but only up to a point. Quick decisions are more likely to lead to a blunder, but after about 10 seconds or so the likelihood of a blunder flattens out. So when players spend more time than this on a move, it is probably because they don’t know what to do.”

“The difficulty of the decision is an important factor, too. More difficult positions are more likely to lead to a blunder.”

“Skill levels have a big impact in reducing the likelihood of a blunder. In general, better players make better decisions.”

And the most surprising: The researchers “have found evidence of an entirely counter-intuitive phenomenon in which skill levels play the opposite role, so that skillful players are more likely to make an error than their lower-ranked counterparts. The team call these ‘skill anomalous positions.'”

In summary

While at this moment most people are planning to adopt formulas to increase their success rate in 2020, I’m mostly focused on avoiding mistakes to reduce my failure rate. Happy New Year!

Posted in Efficacy | Tagged | Leave a comment

On Rules and Probabilities

failureEventually we understand that there are no rules. There are only probabilities. There is a probability to succeed and there is a probability to fail.

No one is able to predict anything. At most one can estimate the chances of success or the risk of failure.

Similarly, it is no use making detailed plans. We can only take preventative measures to lower our risks, or prepare ourselves to the utmost to increase our chances.

We should change our focus. We should stop being obsessive about doing “the right thing”. Because even “the right thing” can fail. And because when we believe we are doing “the right thing” this gives us a false sense of safety.

We should stop trying to make “the right decision”. Because in most cases we never have sufficient information to make “the right decision”. And because we cannot really predict all the possible consequences of our decisions.

In general, critical situations are very complex, and this complexity makes them unique. Since each situation is unique, there is no predefined formula to guarantee a positive outcome. Previous experience is useful but not sufficient.

We should just try to increase our probabilities to succeed. Our we should try to reduce our risks to fail. Because there are simply too many things that we cannot control.

There are no guaranteed results. Never.

Posted in Efficacy | Tagged | 1 Comment

On Keys, Potential and Goals

uncut keyThis is an uncut key. That’s how it comes from the factory.

The uncut key has the potential to open any door. But that is just a potential.

The locksmith cuts the key so that it can open a specific door. When he does this, he is setting a goal for the key.

When we are young, we also start with a huge potential. But that is just a potential.

We also need to prepare to achieve our goals. We need to define which doors we want to open!

Posted in Efficacy | Tagged | 1 Comment

Do we really need New Year’s Resolutions?

new_yearWe are approaching the Jewish New Year. In general, in our New Year’s Resolutions, we always think about what we need to change: for example, we may decide to take an interesting course or start a revolutionary diet.

However, this year I am thinking about what I need to keep: what are the things that are working out for me.

I intend to focus on doing more of what is working, and doing less of what did not work.


In general, I think that people may benefit more from investing in their natural talents instead of constantly trying to develop new skills.

Mastery requires intensive practice for a long period of time, thus it also demands dedication and a clear goal to be achieved.

The modern tendency to try many different paths, always looking for shortcuts, results in the opposite of mastery: shallowness.

Forget New Year’s Resolutions. We probably don’t need to start doing anything new. We just need to persist in what we are doing well.

Shana Tova!

Posted in Efficacy, Jewish Sources | Tagged , | Leave a comment