panosRaptis/DataMining-Project

DataMining-Project

Dataset: "London Bicycle Hires"

Golas of that project was tha analysis of the given the data based on the use of data mining tools, including data understanding, preprocessing and exploration, as well as applying some ML techiniqiues, like clustering analysis and forecasting, to retrieve more insights. For more details of our work, check our report (in Greek).

Notebook's Structure

~/Notebook.ipynb
│
└──Brief Presentation and Data Loading
│
└──Preprocessing
  | Validation - Correction of the Data
  | Remove Outliers
  | Feature Extraction
│
└──Exploratory Data Analysis
   | General Insights
   | Useful Plots and Statistics
│
└──Machine Learning Algorithms
   | Hourly Demand Forecasting
       | DNN
       | Random Forest Regressor
       | GRU
       | LSTM
   | Spatial Clustering Analysis
       | K - means
       | DBSCAN

Libraries Used

pandas, numpy, matplotlib, seaborn, sklearn, keras