Machine Learning

Machine Learning

1. when to use?

   • doing repetitive task
   • getting computers to program themselves
   • writing software is bottleneck

2. supervised

   • classification
   • regression

3. unsupervised

   • clustering

4. Algorithms

   • Prediction
   • classification
   • Clustering
   • Association learning

5. Underfitting

6. overfitting

   • causes poor model perforances
   • Limit overfitting
     • hold back validation dataset
     • use resampling techniques to estimate model accuracy. eg:- K-fold cross validation
   • can lead to poor model performances and most common problem compared to underfitting
   • how well an algorithm performs on unseen data

7. ML activities

   • Preprocess
   • Sampling/ Splitting
     • Split data into training and test data subset
   • Build/Train Model
   • Test Model
   • Measure performances