- Pandas module
- Convert dictionaries to Dataframes
- Slicing dataframes
- Making new columsn in dataframes
- SKLearn and Quandl module
- Get financial and economic datasets using Quandl
- Performing mathematical operations on dataframe columns
- Dataframe functions - .head() .tail() .shift() .fillna() dropna()
- Train, test, predict data using Linear regression or Simple vector machine model
- Features vs labels
- Training and predicting using a model
- Prepare training data and split in 2 parts, ~80% to train ~20% to test [ model_selection.train_test_split() ]
- Define a classifier/model, like LinearRegression, SVM (Simple vector Machine) and then Train the classifier using .fit()
- Test accuracy of the classifier with respect to test data from step 1 [~20% of data]
- Predict - Label = classifier.predict('Features')
- Best fit line and how regression works
- What is slope(m) and intercept(b)
- Linear Regression = mX + b
- What are Squared error?
- Squared error vs Absolute errors
- R-Squared / Coeffcient of determination
- Classification with K-Nearest neighbor (KNN)
-
Euclidean distance
-
Making your own k-NN (k-Nearest Neighbor) algorithm in python
-
Comparing the accuracy and confidence of your algorithm with SKLearn module's neighbors.KNeighborsClassifier()
-
Accuracy vs confidence in k-NN algorithm