Skip to content

Latest commit

 

History

History
47 lines (32 loc) · 1.77 KB

README.md

File metadata and controls

47 lines (32 loc) · 1.77 KB

  • Project Execution By:

    • Danial Soleimany
    • Machine Learning Engineer
    • Student of Artificial Intelligence

  • Project Supervisor:

    • Samaneh Zahedi
    • Bachelor of Radiology from Shiraz University of Medical Sciences
    • Masters Degree in Medical Physics from Jondishapour University of Medical Sciences

🧐 Objectives

  1. Prediction: Predict whether a patient has cervical uterus cancer.
  2. Survival Estimation: Estimate the number of days a patient will survive with cervical cancer.
  3. Machine Learning Application: Apply machine learning algorithms for regression and classification tasks.
  4. Feature Importance: Identify impactful features through feature selection methods.
  5. Dose Rate Investigation: Investigate the relationship between dose rate and survival duration over time.

Challenges

🤒 Challenges

  1. Limited Dataset: Small number of samples challenges achieving high accuracies.
  2. Outliers: Presence of outliers affecting algorithms sensitive to them.
  3. Imbalanced Data: Class imbalance in target variable biasing model learning.
  4. Missing Values: Handling missing values without significant data loss.

Solutions

👩‍⚕️ Solutions

  1. Hyper-parameter Tuning: Prevent overfitting or underfitting using hyper-parameter tuning.
  2. RobustScaler: Mitigate outlier impact by scaling data before model training.
  3. Stratified Sampling: Ensure balanced representation of classes in training and testing sets.
  4. Imputation: Handle missing values with mean and mode imputation to retain sample size.

License

This project is licensed under the MIT License - see the LICENSE file for details.

solutions, and guidance on how to get started with and contribute to the project.