Once when I asked my professor about machine learning he told the following:
Lets say you have a function f(a,b) = ax+by = z . Then there are sophisticated ways to learn a and b from a data set corresponding to [x,y,z]. This is (fitting in) machine learning.
So Here I am writing for someone who wants to use machine learning quickly in python:
First, let me write the general steps:
- Make the data in the right shape
- Create the learning object
- Fit the data using 1 & 2
- Predict on a new test data
from sklearn.ensemble import RandomForestClassifier [X,y] = extract_from_file('data_sheet.txt')#assuming you know #how to write this function rf_classifier_object = RandomForestClassifier(max_depth=2, random_state=0) rf_classifier_object.fit(X,y) #The Learning step rf_classifier_object.predict(X_test)#This will predict the new X_test data
Now you may replace ‘RF’ with any classifier or regressor as you like. Let me know if you have any questions.