Tag: machine learning

Recently I finished reading machine learning for dummies 2nd edition. I did not run any of the code in the book. It was the 2nd time I had read that edition of the book.

I have wanted to do machine learning for years and have nothing to show for it. And I want something to show for it. So hopefully I will start doing some machine learning.

Or maybe I won’t.

to finish and its the Intro to game A.I. and reinforcement learning.

I read through all the other courses and did the exercises. I did not have much of an idea of what to do for the exercises so mostly I uncomment the solution to the problem and copied it to get the answer correct.

I figure I have been on kaggle for over 4 years and I wanted to finally get through all the courses and finally start my own project.

But, the course I really wanted to do with the intro go game A.I and reinforcement learning. I think it would be so fun to write a program that would learn how to play a game on its own. I don’t know how to do that yet.

I though that is what I want to focus on for this year. I am not sure if I will be able to write a probably to play a game but, if its to hard I can always work on some other type of machine learning.

courses. I am on the sql course. I want to do all the courses at once but, I figure if I work on one or two at a time I will eventual get through them all. That is the plan anyway. I don’t know if I will ever be good at machine learning. But, I am giving it a try. I know I need to learn more math.

intermediate machine learning course. I don’t remember when I started on that course. I assume over a year ago.

I think I have down when I need to do for missing values, categorical values. I know a little bit how to use a pipeline, cross validation and XGboost.

I know I need to practice what I learned from the course or I will forget it. I don’t think I will ever be ready to do my own machine learning projects. But, I think getting a machine learning project is what I need to do.

Also the intermediate learning course talked about data leakage. It talked about two kinds of leakage target leakage, and train test spit leakage. At first I did not understand what leakage meant. I think now I have a better idea.

I still have about ten more kaggle tutorials to go through to have completed them all. Probably about 14 to be exact.

I think next I will go through some of the courses I have not finished and work through the feature engineering tutorial and also try a machine learning project. Maybe I will try a competition.

able to remember things better. I read the lesson about how to deal with missing values again. I can either delete the missing values, imputer the values or use an extension to values.

And imputing means that I add values for the missing values. And the extension to imputing is adding values and adding a column that says which values were imputed. It was so hard for me to remember those things when I first read about imputation.

trouble remembering what I had trouble remember the steps I needed to go through to what is it train the model.

I remembered that I needed to load the csv file into a variable. Then after that I could not remember what to do. Now I think I have got the steps down.

Then what I need to do is load the prediction target into a variable. Then I make a list of features. Then I load the features into a variable. Then I take the variable with the features and put it into a data frame.

Then I select the machine learning model I want to use from sci-kit learn, next up is fitting the model, then it time to predict.

Oh, and I forgot that I need to drop missing features from a dataset before I make the variable for the prediction target.

Maybe writing the steps out will help me remember them better.