Fouzanjaved/ImplementationProject

1import pandas as pd2import numpy as np3import tensorflow as tf4from tensorflow.keras.models import load_model5from sklearn.preprocessing import MinMaxScaler6import gradio as gr7import joblib8 9# Load pre-trained model and scaler10model = load_model('diabetes_model.h5')11scaler = joblib.load('scaler.pkl')12 13def predict_diabetes(pregnancies, glucose, insulin, bmi, age):14    """Predict diabetes probability from input features"""15    # Create input array16    input_data = np.array([[pregnancies, glucose, insulin, bmi, age]])17    18    # Scale features19    scaled_data = scaler.transform(input_data)20    21    # Make prediction22    probability = model.predict(scaled_data, verbose=0)[0][0]23    24    # Interpret results25    status = "Diabetic" if probability >= 0.5 else "Not Diabetic"26    confidence = probability if probability >= 0.5 else 1 - probability27    28    # Create explanation29    explanation = f"""30    ### Prediction: {status}  31    Confidence: {confidence:.1%}  32    33    #### Key factors contributing to this prediction:34    - Glucose level: **{'High' if glucose > 140 else 'Normal'}** ({glucose} mg/dL)35    - BMI: **{'Obese' if bmi >= 30 else 'Overweight' if bmi >= 25 else 'Normal'}** ({bmi})36    - Age: {age} years37    - Insulin: {insulin} μU/mL38    - Pregnancies: {pregnancies}39    """40    41    # Create bar chart of feature importance42    features = ['Pregnancies', 'Glucose', 'Insulin', 'BMI', 'Age']43    importance = [0.15, 0.45, 0.10, 0.20, 0.10]  # Example weights44    45    return {46        "probability": float(probability),47        "status": status,48        "explanation": explanation,49        "importance": (features, importance)50    }51 52# Create Gradio interface53inputs = [54    gr.Slider(0, 15, step=1, label="Number of Pregnancies"),55    gr.Slider(50, 200, value=120, label="Glucose Level (mg/dL)"),56    gr.Slider(0, 300, value=80, label="Insulin Level (μU/mL)"),57    gr.Slider(15, 50, value=32, label="BMI (kg/m²)"),58    gr.Slider(20, 100, value=33, label="Age (years)")59]60 61outputs = [62    gr.Label(label="Diabetes Probability"),63    gr.Markdown(label="Explanation"),64    gr.BarPlot(x="Feature", y="Importance", label="Feature Importance")65]66 67title = "Diabetes Prediction App"68description = "Early detection of diabetes using machine learning. Based on research: Khanam, J.J. & Foo, S.Y. (2021)"69article = """70**About this model**:  71- Trained on Pima Indians Diabetes Dataset72- Neural Network with 88.6% accuracy73- Predicts diabetes risk using 5 key health parameters74"""75 76gr.Interface(77    fn=predict_diabetes,78    inputs=inputs,79    outputs=outputs,80    title=title,81    description=description,82    article=article,83    examples=[84        [0, 90, 80, 24, 25],85        [3, 150, 95, 32, 35],86        [6, 180, 150, 38, 45]87    ]88).launch()