Start studying: Saturday, December 7th
Teaching materials, etc .: ・ Miyuki Oshige "Details! Python3 Introductory Note ”(Sotec, 2017): Completed on Thursday, December 19th ・ Progate Python course (5 courses in total): Ends on Saturday, December 21st ・ Andreas C. Müller, Sarah Guido "(Japanese title) Machine learning starting with Python" (O'Reilly Japan, 2017): Completed on Saturday, December 23 ・ Kaggle: Real or Not? NLP with Disaster Tweets: Posted on Saturday, December 28th to Friday, January 3rd Adjustment ・ ** Wes Mckinney "(Japanese title) Introduction to data analysis by Python" (O'Reilly Japan, 2018) **: January 4th (Sat) ~
Challenge Kaggle from the end of last year to yesterday.
Think for yourself, write code, and if there is a function you want to implement, search in books or google to study. When completed, compare it with the kernels of other participants, incorporate any processing that you lacked or have an unprecedented focus, and rewrite the code again. I was able to learn a great deal about machine learning by carrying out a series of processes such as these through actual tasks.
I had been doing trial and error until yesterday to improve my score, but as I tried, I saw various new challenges, so I took a short break here for Kaggle and moved on to inputting knowledge again.
Understanding of various libraries I was keenly aware of my lack of knowledge about libraries such as pandas, numpy, sklearn and matplot. Of course, if you search, you can find as many as you want, and if you look at the kernel, you can understand the functions that are often used, but in addition to the different issues each time, be sure to perform the same processing even if you can pattern it to some extent. I thought it wasn't so often, so I felt I needed to know the big picture at least once about "what each library can do".
Knowledge of statistics Regardless of whether it is visualized with matplot or the like, it is not easy to imagine which number corresponds to where and which number is changed (= cannot be explained to people). I felt that the current situation should be improved.
For the first item, first learn about NumPy and pandas using the books shown in "Materials, etc." at the top of the article. Regarding the second item, we plan to solidify our knowledge by studying while looking ahead to passing Level 2 of the statistical test scheduled to be conducted in June.
p.92 Finish reading up to Chapter 3. Since the tutorial elements such as python built-in functions are strong, the main part according to the original purpose will be after this.
Recommended Posts