Notes / Resources on Machine Learning

Imputation of Missing Values

• Scikit Learn Source

For various reasons, many real world datasets contain missing values, often encoded as blanks, NaNs, or other placeholders. Such datasets are incompatible with scikit-learn estimators which assume that all values in an array are numerical, and that all have and hold meaning.

• You could remove the rows of data or impute them - infer them from known parts of data.
• You can either impute values only by using values from that column - SimpleImputer - or you could use multivariate imputation algorithms to use the entire set of available feature dimensions to estimate the missing values - IterativeImputer
• SimpleImputer allows you to replace missing values with a provided constant value or using statistics from that column (mean, median, or most frequent)
• KNNImputer provides imputation for filling in missing values using the k-Nearest-Neighbors approach.

Cross Validation

• Something you need to review

Precision vs Recall of Classifiers

• Something you need to review

Bagging and Pasting

• Something you need to review

Kernel Trick

• Something you need to review

Random Forests

• Something you need to review