Here are my/my_team solutions for the Kaggle competitions I've taken part in, showing my approaches and insights. Details of each solution are inside the folder.
Task: Create a model to predict how well one Monte-Carlo tree search (MCTS) variant will do against another in a given game, based on a list of features describing the game.
Rank: 9th out of 1610 teams.
Type: Tabular Modeling.
Competition URL: https://www.kaggle.com/competitions/um-game-playing-strength-of-mcts-variants
Solution Write-up: https://www.kaggle.com/competitions/um-game-playing-strength-of-mcts-variants/discussion/549624
Task: Predict which responses users will prefer in a head-to-head battle between chatbots powered by large language models (LLMs).
Rank: 19th out of 1688 teams.
Type: Natural Language Processing, Large Language Models (LLMs), Classification.
Competition URL: https://www.kaggle.com/competitions/lmsys-chatbot-arena
Solution Write-up: https://www.kaggle.com/competitions/lmsys-chatbot-arena/discussion/528288
Task: Build a model that evaluates how well a student represents the main idea and details of a source text, as well as the clarity, precision, and fluency of the language used in the summary.
Rank: 27th out of 2064 teams.
Type: Natural Language Processing, Classification.
Competition URL: https://www.kaggle.com/competitions/commonlit-evaluate-student-summaries
Solution Write-up: https://www.kaggle.com/competitions/commonlit-evaluate-student-summaries/discussion/446542
Task: Develop a model based on a large dataset of keystroke logs that have captured writing process features to predict overall writing quality.
Rank: 36th out of 1876 teams.
Type: Tabular Modeling, Natural Language Processing.
Competition URL: https://www.kaggle.com/competitions/linking-writing-processes-to-writing-quality
Solution Write-up: https://www.kaggle.com/competitions/linking-writing-processes-to-writing-quality/discussion/466839
Task: Build a multi-objective recommender system based on previous events in a user session to predict e-commerce clicks, cart additions, and orders.
Rank: 56th out of 2574 teams.
Type: Recommendations Systems, Tabular Modeling.
Competition URL: https://www.kaggle.com/competitions/otto-recommender-system
Solution Write-up: Code only (Summary inside the folder's readme).
Task: Develop a model to construct precise 3D maps (spatial representations) using sets of images in diverse scenarios and environments regardless of the source domain—images taken from drones, amidst dense forests, during nighttime, or any of the six problem categories.
Rank: 71st out of 929 teams.
Type: Computer Vision, Images Reterival and Matching.
Competition URL: https://www.kaggle.com/competitions/image-matching-challenge-2024
Solution Write-up: Code only (Summary inside the folder's readme).
Task: Develop a model based on data from the "IceCube" detector, which observes the cosmos from deep within the South Pole ice, to predict a neutrino particle’s direction.
Rank: 78th out of 812 teams.
Type: Tabular Modeling, Astronomy/Physics.
Competition URL: https://www.kaggle.com/competitions/icecube-neutrinos-in-deep-ice
Solution Write-up: Code only (Summary inside the folder's readme).
Task: Create a model to process birds audio data and recognize the species by their calls.
Rank: 99th out of 1189 teams.
Type: Audio/Waves, Computer Vision.
Competition URL: https://www.kaggle.com/competitions/birdclef-2023
Solution Write-up: Code only (Summary inside the folder's readme).
Task: Develop a model that match the radiologists' performance in detecting and localizing fractures to the seven vertebrae that comprise the cervical spine.
Rank: 100th out of 883 teams.
Type: Computer Vision, Medical Images.
Competition URL: https://www.kaggle.com/competitions/rsna-2022-cervical-spine-fracture-detection
Solution Write-up: Code only (Summary inside the folder's readme).
Task: Create a model that can be used to aid in the detection and classification of degenerative spine conditions using lumbar spine MR images.
Rank: 104th out of 1874 teams.
Type: Computer Vision, 3D Medical Images.
Competition URL: https://www.kaggle.com/competitions/rsna-2024-lumbar-spine-degenerative-classification
Solution Write-up: Code only (Summary inside the folder's readme).
Task: Train a model to score student essays.
Rank: 188th out of 2706 teams.
Type: Natural Language Processing, Classification.
Competition URL: https://www.kaggle.com/competitions/learning-agency-lab-automated-essay-scoring-2
Solution Write-up: Code only (Summary inside the folder's readme).
Task: Develop a model to assess the language proficiency of 8th-12th grade English Language Learners (ELLs).
Rank: 223th out of 2654 teams.
Type: Natural Language Processing, Classification.
Competition URL: https://www.kaggle.com/competitions/feedback-prize-english-language-learning
Solution Write-up: Code only (Summary inside the folder's readme).
Task: Solve 27 hand-crafted machine learning security challenges to find flags, solve puzzles, and gain hands-on experience with concepts of AI security and safety.
Rank: 166th out of 1345 teams.
Type: AI Security, Adversarial Learning.
Competition URL: https://www.kaggle.com/competitions/ai-village-capture-the-flag-defcon31
Solution Write-up: https://www.kaggle.com/competitions/ai-village-capture-the-flag-defcon31/discussion/454366
Task: Develop a model trained on one of the largest open datasets of game logs to predict student performance during game-based learning in real-time.
Rank: 276th out of 2051 teams.
Type: Tabular Modeling, Natural Language Processing.
Competition URL: https://www.kaggle.com/competitions/predict-student-performance-from-game-play
Solution Write-up: Code only (Summary inside the folder's readme).
Task: Develop a model that can predict the ranking of protein thermostability (as measured by melting point, tm) after single-point amino acid mutation and deletion.
Rank: 290th out of 2483 teams.
Type: Biology, Sequence Data, Computer Vision, Tabular Modeling.
Competition URL: https://www.kaggle.com/competitions/novozymes-enzyme-stability-prediction
Solution Write-up: Code only (Summary inside the folder's readme).
Task: create a model trained on 3D Hierarchical Phase-Contrast Tomography (HiP-CT) data from human kidneys to segment blood vessels.
Type: Computer Vision, Segmentation, 3D Medical Images.
Competition URL: https://www.kaggle.com/competitions/blood-vessel-segmentation
Solution Write-up: Code only (Summary inside the folder's readme).
Task: Create a model that can solve tricky math problems written in LaTeX format.
Type: Natural Language Processing, Large Language Models (LLMs).
Competition URL: https://www.kaggle.com/competitions/ai-mathematical-olympiad-prize
Solution Write-up: Code only (Summary inside the folder's readme).
Task: Develop a model trained on electroencephalography (EEG) signals recorded from critically ill hospital patients to classify seizures and other types of harmful brain activity.
Type: Signal Processing, Waves Data, Computer Vision.
Competition URL: https://www.kaggle.com/competitions/hms-harmful-brain-activity-classification
Solution Write-up: Code only (Summary inside the folder's readme).
Task: Develop a classification model that accurately categorize the Arabic poems based on their era.
Type: Natural Language Processing, Classification.
Competition URL: https://www.kaggle.com/competitions/arabic-poem-classification
Solution Write-up: https://www.kaggle.com/competitions/arabic-poem-classification/discussion/483099