Skip to content

Latest commit

 

History

History
72 lines (65 loc) · 2.51 KB

README.md

File metadata and controls

72 lines (65 loc) · 2.51 KB
Raw

JosuaLearnToMakePrediction

This research or project is a part that I also use as a final project (dissertation). There are two documents with modeling methods using RNN and LSTM.

Predictions of Tourist Visits at Taman Nasional Gunung Merbabu Using the RNN and LSTM

Some methods that I have used to make predictions (Regression Problems)

  • Univariate Time Series Analysis
  • Resampling Using Cubic Spline
  • Prpocessing data (Stasionary, Normalization, Supervised learning)
  • Modeling with RNN and LSTM
  • Evaluation (RMSE)

First Folder, 01-TugasAkhirEnv-RNN

  • Time series analysis (Stasionerity)
  • Upsampling from monthly to weekly (Interpolation)
  • Upsampling from monthly to daily (Interpolation)
  • Predict With monthly data (01-XX)
  • Predict With weekly data (02-XX)
  • Predict With daily data (03-XX)
  • 3 dataset (CSV)

First Folder, 02-TugasAkhirEnv-LSTM

  • Time series analysis (Stasionerity)
  • Upsampling from monthly to weekly (Interpolation)
  • Upsampling from monthly to daily (Interpolation)
  • Predict With monthly data (01-XX)
  • Predict With weekly data (02-XX)
  • Predict With daily data (03-XX)
  • 3 dataset (CSV)

File name description

  • XX=01 One feature predict for one label

  • XX=02 Many feature predict for one label

  • XX=03 Many feature predict for many label

  • M11 Monthly dataset, One feature, One label

  • MM1 Monthly dataset, Many feature, One label

  • MMM Monthly dataset, Many feature, Manay Label

  • W11 Weekly dataset, One feature, One label

  • WM1 Weekly dataset, Many feature, One label

  • WMM Weekly dataset, Many feature, Manay Label

  • D11 Dailly dataset, One feature, One label

  • DM1 Daily dataset, Many feature, One label

  • DMM Daily dataset, Many feature, Manay Label

  • (.h5) Saved model

  • (.png) Layer visualization

Tools and Environment

  1. Personal Computer (PC) MSI Gaming Notebook GL62 7RD.
  2. Intel(R) Core i7-7700HQ CPU @ 2.80GHz
  3. Nvidia GeForce(R) 1050 GTX Series 2GB GDDR5
  4. Memory DDR4-2400 8GB
  5. Windows 10 Professional x64
  6. Anaconda 3 Version 2019.07
  7. Visual Studio Community 2017 Version 15.9.15
  8. Nvidia Driver Geforce GTX 1050 Version 441.41
  9. Nvidia CUDA Toolkit Version 10.0.130
  10. Nvidia CuDNN Version 7.0
  11. Conda Virtual Environment Version 4.7.10
  12. Virtual Environment Set Up:
    • Python Version 3.6.9
    • Juptyter Notebook Version 6.0.1
    • Numpy Version 1.16.5
    • Scikit-Learn Version 0.21.3
    • Scipy Version 1.3.1
    • Pandas Version 0.25.3
    • Statsmodels Version 0.10.1
    • Matplotlib Version 3.1.1
    • Tensorflow-GPU Version 2.0.0
    • Keras Version 2.3.1