Open elucidator8918 opened 1 year ago
@elucidator8918 I want to work on this project, I am a girls script summer of code aspirant. Please assign it to me.
@elucidator8918 - you can go ahead! We are assigning you 21 days for this project, after which it will be assigned to someone else if not completed. All the best!
Name the file as: algorithm_dataset.ipynb
and link it in the readme of the labeled directory as algorithm - dataset
.
@varsha089 - since we are following the first-come-first-serve policy, we will not be able to assign you this issue. However, you can create another issue on the similar lines.
💥 Proposal
Hello, I am Siddhant Dutta - I am a GSSOC'23 Contributor. This is my resume - Resume
Introduction
The Price Suggestion Model is an open-source project aimed at predicting the prices of products listed on online marketplaces like Amazon. The goal is to provide sellers with accurate price suggestions for their products based on various attributes such as brand, category, description, name, shipping, and item condition. This project utilizes machine learning algorithms, including LGBM and Neural Networks, to predict prices and employs techniques like Bayesian search and ensemble methods to optimize model performance.
Objectives
1) Develop a machine learning model capable of predicting prices for products listed on online marketplaces. 2) Implement various algorithms, including LGBM and Neural Networks, to achieve accurate price predictions. 3) Utilize techniques such as Bayesian search and ensemble methods to optimize the performance of the price prediction model. 4) Evaluate the model's performance using metrics such as Mean Squared Error (MSE) and Root Mean Squared Logarithmic Error (RMSLE). 5) Provide an open-source solution that can be used by sellers on marketplaces like Amazon to improve their pricing strategies.
Deliverables:
The project will deliver the following components: 1) Source code of the price suggestion model implemented in Python. 2) Preprocessing scripts for cleaning and transforming the dataset. 3) Jupyter notebooks or Python scripts showcasing the implementation of various machine learning algorithms. 4) A trained model file that can be used for price prediction as well as the h5 weight file of the final deep learning model. 5) Documentation detailing the project's methodology, including preprocessing steps, algorithm selection, hyperparameter tuning, and evaluation metrics.
Methodology:
1) Obtain a suitable dataset from platforms like Kaggle, containing product attributes and their corresponding prices. 2) Preprocess the dataset, including cleaning and transforming the data, handling missing values, and encoding categorical variables. 3) Split the dataset into training, cross-validation & testing sets to train and evaluate the machine learning models. 4) Implement machine learning algorithms, including LGBM and Neural Networks, to build the price prediction model. 5) Utilize techniques like Bayesian search and ensemble methods to optimize the model's hyperparameters and overall performance. 6) Evaluate the model's performance using custom loss functions & metrics such as MSE and RMSLE, comparing the predicted prices with the actual prices in the test dataset. 7) Iterate and refine the model based on the evaluation results to improve accuracy.
Timeline
EDA, Data Cleaning & preprocessing: 1 week Feature Engineering Model implementation: 2 weeks Optimization & Evaluation and refinement: 1 week Documentation and finalization: 1 week