CancerDetection

Goal:

1. Build a classifier that can distinguish people with cancer and people without cancer in each age group (Young and Mature groups).
2. Discover the meaningful gene structures (signatures) that have an impact to the cancer detection

Current Approach: HSIC Lasso + Bayesian Optimization + SVM

Usage of extrFeat.py: extract a small number of features by using HSIC Lasso

# python <script> <original_data_filepath> <label_name>
#   label_name = "Diagnosis", "Class"

Usage of Bo.py: apply Bayesian Optimization to tune a classification model

python <script_name> <classifier_name> <label_name> <original_data_filepath>
#    classifier_name  = "SVM", "RF"
#    label_name =  "Diagnosis", "Class"
#    original_data_filepath:
#       e.g., Sample_Master_Source_and_Platform_batch_removed_Extra_Sample_Info.exported_for_AI.txt
#
# The script should be in the same directory level as directories where the extracted features files locate

Clustering Analysis On Extracted Features: check the notebook postAnalysis.ipynb