bandicam.2022-08-21.20-55-14-815.online-video-cutter.com.mp4
This project is an online analytical dashboard for analyzing TikTok data, developed as part of the Software Engineering Course at the University of Zanjan.
The goal of this project is to analyze trending tiktoks data using machine learning techniques. By leveraging a dataset extracted from TikTok and performing exploratory data analysis (EDA) tasks, the project aims to preprocess the data and train a regression model. The selected model for training is the Extra Trees Regressor, chosen based on my experience with different models.
The latest TikTok dataset from Kaggle was gathered for this project. Raw data was subjected to EDA tasks to gain insights and understand the underlying patterns and characteristics of the dataset. This enabled effective preprocessing of the data for subsequent steps.
- Identified the problem.
- Extracted use cases and requirements.
- Gathered the latest TikTok dataset from Kaggle.
- Conducted exploratory data analysis (EDA) on the raw data.
- Performed necessary data preprocessing steps.
- Determined that the task is regression based on the problem definition.
- Considered the data format and volume to choose the Extra Trees Regressor model.
- Leveraged experience working with different models to make an informed selection.
- Trained the Extra Trees Regressor model on the preprocessed dataset.
- Evaluated the model's performance and adjusted parameters as needed.
- Utilized Streamlit to deploy the web application on the internet.
- Provided an interactive interface for users to detect trends using the trained model.
I would like to express my gratitude to Nicholas Renotte for inspiring this project. His initial idea served as a foundation for my own implementation, allowing me to enhance the web app further.