Used Car Price Regression with Neural Networks

The Challenge

The price of a used car is determined by numerous factors, such as its brand, model, age, mileage, and condition. The challenge was to build a machine learning model that could analyze a dataset of used car listings from the OLX platform and accurately predict a car's price based on its features.

The Solution

As a project for the Google Bangkit Academy, I developed a regression model to tackle this prediction task. The core of the solution was a Neural Network designed specifically for regression.

The Machine Learning Workflow

• Data Analysis & Preparation: I used the Pandas and Numpy libraries in Python to load, clean, and analyze the OLX dataset. This involved handling missing values, encoding categorical features (like car brand and model), and normalizing numerical data.
• Model Building with TensorFlow: I built the neural network model using TensorFlow. This involved defining the architecture of the network, including the number of layers and neurons, and choosing the appropriate activation functions and optimizer for a regression problem.
• Training & Evaluation: The model was trained on the prepared dataset, and its performance was evaluated using standard regression metrics like Mean Absolute Error (MAE) and Mean Squared Error (MSE) to determine its accuracy in predicting prices.

This project demonstrated my ability to apply neural networks to a classic regression problem, covering the essential steps of data preparation, model building, and evaluation.