Predictive Modeling for Breast Cancer Treatment by El Bacha IliasPredictive Modeling for Breast Cancer Treatment by El Bacha Ilias

Predictive Modeling for Breast Cancer Treatment

El Bacha Ilias

Completed work

Data Scientist

pandas

scikit-learn

TensorFlow

Healthcare

Neoadjuvant Therapy Classification Project

Project Overview

This project focuses on predicting the response to neoadjuvant therapy in cancer patients using various machine learning models. The dataset includes clinical and histopathological features of patients. The project includes data processing, feature engineering, model training, and evaluation.

Machine Learning Models and evaluation

Data Description

The dataset used for this project is Neoadj_base_luminaux - anonymise.xls. It contains clinical data and histopathological information for patients undergoing neoadjuvant therapy.

Columns: The dataset has multiple columns including patient identifiers, clinical data, histopathological features, and outcome variables.

Setup

Requirements

To run this project, you'll need the following Python packages:

pandas

numpy

matplotlib

seaborn

scikit-learn

imbalanced-learn

xgboost

tensorflow

You can install the required packages using pip:

pip install pandas numpy matplotlib seaborn scikit-learn imbalanced-learn xgboost tensorflow

File Structure

/path/to/project
│
├── Neoadj_base_luminaux - anonymise.xls # Raw dataset
├── perso_mec_imp.csv # tailored imputed dataset
├── data_preprocessing.py # Data processing and feature engineering script
├── model_building.py # Machine learning model training and evaluation script
├── neural_network.py # Neural network model implementation and training script
├── README.md # This README file
└── requirements.txt # List of Python dependencies

Data Preparation

In the data_preprocessing.py script, the following steps are performed:

Loading Data: Load the dataset and additional features.

Data Cleaning: Drop irrelevant columns and encode categorical variables.

Feature Engineering: Merge features from different sources and select relevant features.

Handling Imbalanced Data: Apply SMOTE to balance the classes.

# Load the dataset
df = pd.read_excel('/path/to/Neoadj_base_luminaux - anonymise.xls')

# Data Cleaning and Feature Selection
df1 = df.drop([...], axis=1)  # Drop irrelevant columns

# Encoding categorical variables
sata_n_map = {'A': 0, 'B': 1, 'C': 2, 'D': 3}
df1['sata_n'] = df1['sata_n'].replace(sata_n_map)

# Filtering data based on 'chim_typ'
filtered_df = df1.loc[(df1['chim_typ'].isin([1, 2, 5]))]
grouped_df = filtered_df.groupby('anap_kc').apply(lambda x: x.reset_index(drop=True))

# Load additional features
X_df = pd.read_csv('/path/to/perso_mec_imp.csv')

Feature Selection

Different feature selection methods are employed:

ANOVA: anova_top8 function selects features based on ANOVA test results.

Gini Index: GiniRF_top8 function selects features based on the Gini index from a Random Forest classifier.

SVM ANOVA: svm_anova function selects features relevant for SVM classification.

Machine Learning Models and evaluation

Various classifiers are trained and evaluated, including:

Random Forest Classifier

Support Vector Classifier

Logistic Regression

Gradient Boosting Classifier

XGBoost Classifier

The performance of the models is evaluated using cross-validation and metrics such as accuracy, specificity, sensitivity, and confusion matrices.

Like this project

Completed work

Posted Jun 15, 2025

Predicted response to neoadjuvant therapy using machine learning models.

Likes

Views

Predictive Modeling for Breast Cancer Treatment

Neoadjuvant Therapy Classification Project

Project Overview

Table of Contents

Data Description

Setup

Requirements

File Structure

Data Preparation

Feature Selection

Machine Learning Models and evaluation

Predictive Modeling for Breast Cancer Treatment

Neoadjuvant Therapy Classification Project

Project Overview

Table of Contents

Data Description

Setup

Requirements

File Structure

Data Preparation

Feature Selection

Machine Learning Models and evaluation