This course emphasizes support vector machine (SVM) algorithms and neural networks to improve predictive accuracy, reduce bias and lower variance in real world applications. Case studies compare the advantages and disadvantages of machine learning and neural network models with traditional regression models. The iterative nature of this course provides review and reinforcement of common practices of the predictive modeling cycle.
- Apply SVM's for estimating binary and categorical outcomes.
- Choose suitable SVM algorithms including stochastic gradient descent, ridge, lasso and elastic net.
- Analyze and test different SVM development using bagging to reduce variance, stacking to improve accuracy and boosting to reduce bias.
- Apply common techniques for selecting estimators.
- Explain how a fundamental neural network is constructed and tuned.
- Develop a convolutional neural network models for effective dimension reduction.
- Include principal component analysis within predictive modeling for dimension reduction.
- Combine regression, SVM and neural network algorithms for optimizing predictive models.
- Incorporate code libraries to complete data management and predictive modeling tasks.
|
Week |
Date |
Topics |
|
|
1 |
08-Jan-25 |
Pipelines Decision trees |
|
|
2 |
15-Jan-25 |
Advanced tree-based algorithms |
|
|
3 |
22-Jan-25 |
Ridge, Lasso, and Elastic Net Support vector machines SHAP Ensembles & Stacking |
Assignment 1 (Due in week 10) |
|
4 |
29-Jan-25 |
Gradient Descent Artificial neural network introduction |
|
|
5 |
05-Feb-25 |
Artificial neural network continued |
|
|
6 |
12-Feb-25 |
Artificial neural network continued |
|
|
7 |
19-Feb-25 |
Artificial neural network continued RNN, CNN and LSTM introduction |
|
|
8 |
26-Feb-25 |
Midterms |
|
|
9 |
05-Mar-25 |
Object detection |
Assignment 2 (Due in week 14) |
|
10 |
12-Mar-25 |
Automated machine learning Instructional Assessments Online |
|
|
19-Mar-25 |
Reading Week |
||
|
11 |
26-Mar-25 |
PySpark introduction |