Here is the compilation of the Data Science, AI, Data Mining, and Cloud Computing
-
Data Science
- Python
- R
- SQL
- Machine Learning (Regression, Clustering)
- Tableau
- Jupyter Notebook
-
Artificial Intelligence (Python)
- Search Algorithms
- Uniform Cost Search and A-Star (A*)
- Markows Decision Process (MDP)
- Search Algorithms
-
Data Mining
- Python
- Used Cross Validation method for classifiers
- ML (Clustering)
- Python
Analysed the impact using data provided by the Government and other Health Authorities to find the trend and also to predict the trend.
- Conducted an investigation into influenza trends before and during the COVID‑19 pandemic in Australia
- Analyzed the impact of COVID‑19 on influenza, including identifying the most at‑risk demographics based on sex and age
- Concluded with insights on whether there should still be a concern for influenza in the presence of COVID‑19, utilizing tools like Python(Pandas, Seaborn, Matplotlib), R, MySQL, Jupyter Notebook, and Tableau for analysis and visualization.
This project is a web-based application that allows you to chat with multiple PDF documents. It uses natural language processing and machine learning techniques to extract text from PDFs, create embeddings, and provide conversational responses to user questions.
- Upload and analyze multiple PDF documents and Extract text from PDF documents.
- Split large text into manageable chunks along with Providing conversational responses to user questions.
The DragonGame AI Environment, "Untitled Dragon Game" or simply DragonGame, is a 2.5D platformer game inspired by the "Spyro the Dragon" series. In DragonGame, the player's objective is to collect all the gems in each level and reach the exit portal while navigating a 2D grid of tiles. The game involves a jump-and-glide movement mechanic, and the player must avoid landing on lava tiles. And implemented the AI agent to find the optimal solutions and clear the stage using search algorithms with the decision process(MDP)
- Developed intelligent agent for a 2.5D platformer game inspired by ”Spyro the Dragon,” focusing on collecting gems and reaching an exit portal while navigating a dynamic and hazard‑filled environment
- Utilized search algorithms(UCS, A*) in the first part of the project to find optimal paths for the player character (the dragon) while minimizing action costs
- In the second part, Markov Decision Process (MDP) algorithms to create policies for the dragon’s actions, considering uncertain outcomes. This project showcases expertise in game AI, Pathfinding, and decision‑making, enhancing your portfolio.
Still in Progress...
- Developed a robust classifier for a challenging CIFAR-10-derived dataset, addressing real‑world complexities such as missing values, diverse data scales, and outliers.
- Utilized Pre‑processing techniques, including imputation, normalization, and outlier detection, selected based on their impact on classification performance during cross‑validation.
- Applied four classification methods (Decision Tree, Random Forest, k-nearest Neighbor, and Naïve Bayes) with model selection, Hyper‑parameter tuning, and Ensemble learning achieving accurate categorization of data points into ten distinct classes.
This project is licensed under the MIT License - see the LICENSE.md file for details