Statistical_analysis

Statistical analysis using only C++

• Matrix calculation
• Linear regression
• Multiple regression
• Sparse regression
• Logistic regression
• Nonlinear regression
• TimeSeries Regression
• Nonlinear classification
• TimeSeries classification
• Principal component analysis
• Independent component analysis
• Eigenvalue analysis
• Statistical causal search(LiNGAM)

---

The current aim is to create a non-GPL C++ implementation of a linear non-Gaussian aperiodic model (LiNGAM) which is one of the statistical causality search.
LiNGAM

Please wait.
Execution efficiency and results are still bad.
We are currently making modifications and changes.
GPU support is experimental so performance is not good.
The directory structure and file name will probably change in the future.

In reality, Since the GPU calculation is not an experimental level and the mounting method is very bad, the calculation speed is slower than the calculation by the CPU. That is the worst.

progress

Update 2018/11/27

---

We use the following code according to the license.

hungarian-algorithm

Copyright (c) 2014, Markus Buehren
Copyright (c) 2018, mcximing
MIT License Copyright (c) 2018 Sanaxen(Made a single header)

Linear Regression

• Lasso
• Ridge
• Elastic Net
  MIT License Copyright (c) 2017 Liu MIT License Copyright (c) 2018 Sanaxen

logistic regression

• logistic regression

BSD 3-Clause License Copyright (c) 2007-2019 The LIBLINEAR Project

CLAPCK

modified BSD license
Copyright (c) 1992-2013
The University of Tennessee and The University of Tennessee Research Foundation. All rights reserved.
Copyright (c) 2000-2013
The University of California Berkeley. All rights reserved.
Copyright (c) 2006-2013
The University of Colorado Denver. All rights reserved.

MKL(optional)

Copyright (c) 2017 Intel Corporation.

CSVparser

MIT License Copyright (c) 2017 Romain Sylvian
MIT License Copyright (c) 2018 Sanaxen(Made a single header)

tiny-dnn

BSD 3-Clause License Copyright (c) 2013, Taiga Nomi
Used for nonlinear regression and nonlinear time series prediction