[PYTHON] For those who will take AI, ML, and data scientist school courses from now on

** Nice to meet you, I'm Momo-chan's dad. ** **

It seems that the title suddenly seems to be great, but I have taught more than 200 people at a certain school so far. As a mentor widely, including Python, AI (not ML, but ...), DataScience, Blockchain (Solidity) At first I was changing jobs with the enthusiasm that I could increase the number of colleagues who will work in the same industry instead of teachers, but after the job was decided As a mentor between jobs (in order to be able to work for newcomers, not teachers or teachers) I've been active (such as consulting with other things). This can change where Japan is said to be late! !! I was enthusiastic and put off my work (although it will be all night ... (laughs)).

** I will stop mentoring at school in 2021 **

Reason ** 1. I learned that the person who took the course could not change jobs or get a job = I do not understand, I can not use it ** ** 2. The mentor can't even build the environment (recreate it because the environment is broken = all the results so far will disappear) **

** The school operator will say, "Isn't it out?" "Isn't it done?" ** ** Facts are facts. Realization cannot be changed. So far, I have received many cases and help. Don't you know that? When There were many times when I was stunned. I'm the type who talks to people I meet (on the contrary, I'm not good at first-timers, but I'm quick to get to know them. Is it your job to be tampered with? (Because it's like (laughs)), I think too much about the other person's future and say too harsh things If the other person is suitable, I have adjusted it appropriately (need to do it seriously to the appropriate person?).

*** Why did you know? Ironically, this was feedback from the students (the school wouldn't have predicted it either). *** *** It seems that there was an opportunity for a professional from the same school who studied data science at the same school to have an interview. (The interviewer was my student, and I was too far from mathematics to come to study data science in Python.) The interview went on, and he talked about pays statistics, simulations, and so on. But that person said:

** "That's not a data scientist" **

This is the end of the interview. This Even if you can program with Python or solve the problems of data science theory (Bayesian statistics etc.) early Same as the exam, the answer is correct. But the people you need in the field are those who understand the data. At first, I think there are various things, but I want someone who has that attitude. "A person who can see the data, formulate and test hypotheses about the theory up to the goal, and explain the results theoretically." So, if you just give an answer like 1 + 1 = 2, a PC is enough. (Actually, there are many people who make a mistake and only calculate on a PC.)

There is no need to separate science and humanities, and proper education is provided.

There are also doctors among the students. I am taking time to participate while I am very busy trying to make use of it for medical treatment. I feel that not only doctors but also students can be roughly divided into two types. (1) I am very enthusiastic about studying, but I do not understand the program well, so submission of assignments is slow. (2) Finish all the assignments early and pass all the assignments at the start of the mentor (although there is a question when you look at it later)

We ask these two students, "As a result of machine learning, we were able to analyze 600 images in 17 seconds with 86% accuracy. Can we use it for work?" How would you answer? "Because it's medical, it's useless without 99% accuracy." → That's right. But there is no reason to have it to be 99%. "If a person does it, it can't be done in 17 seconds, so I'll think about what to do with the rest." → If a person does it, can 86% = 516 sheets be correctly identified in 17 seconds? (If possible, it will be TV news (laughs))

** All you need is the latter person. A person who can verify the results of machine learning by mobilizing common sense, industry knowledge, and other necessary knowledge. ** **

Since the IT industry is long, I have created many programs. I have seen it. I have reviewed it. Recently, unit test level defects often occur in production. That's right, I'm only thinking about the immediate future. The IT industry may be fascinating. I'm fine with death marches (although most of them are caused by incompetent managers these days). If you want to become an engineer, please learn programming. People who can make programs are normal. Combine various programs I think it would be nice to have a school that teaches you that you can achieve your goals by the deadline by debugging and adding functions. (Which one? I haven't been there ... w)

・ People who can program, not programs ・ Person who can explain the basis of numerical value I would like to take on the challenge of nurturing. At the beginning of the year 2021, is it all right with the old moyamoya? I tried to make an article to make it refreshing.

Recommended Posts

For those who will take AI, ML, and data scientist school courses from now on
Commentary on unbiasedness and consistency for those who don't like it
A story of a person who started aiming for data scientist from a beginner
AWS ~ For those who will use it ~
Data Science Virtual Machines is the best environment for data analysis from now on!