Embedded Files
The challenge for the project is based on a dataset obtained from Pelontar XI basketball association.
Group members: Malik and Marie
This group has dedicated their project for 3-points, 2-points and attack profiling analysis in solving the challenge using the basketball team's dataset. They used clustering to group the Pelontar XI performances based on the opponents they faced during the season.They used RapidMiner and Python. This type of analysis is a team based-analysis which is different from a player based-analysis in a sense that it allows coaching staff to create new trainings and new strategies based on teamwork rather than specified training for each player, or new roles for each player. The team also developed attack profiling analysis. They also built solution using Pokemon Go dataset. They utilized the attack and defense stat for the digital game's analysis. 3D clustering of attack profiles was also developed.
These girls built a dashboard for the visualisation of each of the basketball player's achievements in each game. They also developed simulator and performed experiments to compare the performance of various machine learning algorithms developed through various split ratio to predict the chances of winning or losing a game. They also built the predictor of diabetes using Flask.
Group members: Raja Nur Amanina and Nurrainy
The objectives of this group for the basketball analysis are
To predict the team's factors (such as offensive) that affect the game whether win or lose to make the coach can adjust the training program for the team.
To predict the performance / ranking of player from PXI Basketball Association to easier the PXI manager to plan a suitable training package to improve the weaknesses of players.
They are interested about properties in Malaysia. Therefore, they use the property dataset from kaggle.com to:
Analyse and explore the various characteristics of properties in data set.
Identify how properties vary between location.
Analyse the factor which affect the price of property the most.
Group members: Mow, Chua and Azizah
Group members: Eng Tian, Wan Mardhiah, Ain
This group focuses on regression approach to make the simulator for prediction of whether the basketball team will win or lose, and the predict performance of the player.
This group conducted data mining experiment for predicting the outcome of a match and grouping players according to their profile. They used RapidMiner for building the classifier and wrote Python codes for clustering application. Further, they used a credit card dataset and discovered that purchases and payments are also determinant factor to predict if a credit card holder is a frequent buyer. They also performed clustering to understand the behaviour of credit cards holders.
Group members: Florian, Paul, Elliot, Pierre
Group members: Shahirah and Aishah
This group has come out with a dashboard using the RapidMiner Server for the basketball data exploration and analysis. The dashboard visualizes the performance of each player in each game. The dashboard allows the basketball team to analyse the top players in each game, and across games. Top players in each skill e.g., assist, rebound, turnover, block and steal are also visualized.
Using the second dataset, an income prediction model is built demonstrated through a flask application.
This group consists of 3 clever and committed students from China. They have developed a Python application for the basketball player ability analysis (e.g., offensive, defensive). Visualization techniques such as scatter plot and radar were used. Clustering analysis was used to group players with similar abilities. Various machine learning models such as logistic regression model, decision tree model, Gaussian Bayes, KNN model, and SVM model were developed for the ability analysis. Correlation analysis and association rules were also applied. The second dataset used is heart disease and solutions were built with RapidMiner.
Group members: Chen Yaping, Ni Xiaoyan, Li Jiayu
This group uses the dataset on the record of master study application to develop a prediction model on the percentage of chance for the candidates to apply master's degree. They developed clustering model to group player in their categories based on the player’s performance. This is so that to the coach will have statistical information to monitor each player’s performances and the players will know their potential to play in future and to improvise their skills in future games. The tendency to win the games in future will be increased.
Group members: Thassh, Kirtana, Prisha
This group conducted data mining experiment to evaluate the performance of using player information to predict winning or losing. They also performed clustering. The mobile price dataset has been used to demonstrate their skills in predictive modeling.
Group members; Zheng Tianyi, Yu Jingjing
This all male group focuses on the winning prediction, besides analysis of player's strength. RapidMiner server is used to build association rules, neural network and clustering. The second dataset is from NBA.
Group members: Daniel, Imran and Alif
Azhan and Syatir came out with data analysis on each of the game using PowerBI. They predict WINING or LOSING using K-Nearest Neighbours, linear regression, logistic regression, neural network, decision tree and random forest model. For the second dataset they build dashboard for EPL.
data mining slides latest hafiz.pptxDennis applied Flask framework and other Python libraries on the IMDB and Iris dataset to build models such as Logistic Regression, Linear Discriminant Analysis, K-Nearest Neighbors, Decision Tree Classifier, Gaussian Naive Bayes and Support Vector Machine.
Google Sites
Report abuse