Machine Learning: process of turning data into actionable knowledge for task support and decision making

Define problem → Model data → Evaluate → Deploy

1. Basic Math for computational data analysis
2. Unsupervised learning(EDA: exclusive data analysis) for data exploration
3. Supervised learning for predictive analysis

• Xnxd is our dataset
• Y → labels or target values
• Point rows: instance (sample data point)
• Columns: dimensions(in machine learning, we can have millions of dimensions), features, attributes, variables

Entire dataset = Training dataset

These matrices are the raw data → we create a ML model using this training dataset

• supervised learning: there will be the random test data point to figure out the output label based on the model
• unsupervised learning: we don’t have Y but X, so we cluster the data points in different buckets based on similarities;
• semi-superivsed learning: possible to utilize unsupervised learning to improve supervised learning

Unsupervised Learning:

Clustering Analysis

K-means

Gaussian mixture model

Hierarchical clustering

Density-based clustering

Evaluation of clustering algorithms

Dimension Reduction: reduce the dimensions of the dataset b/c data collection with multiple features cost heavily