Merge "[DOCS] This patch adds documentation."

[thoth.git] / docs / models / models.rst

diff --git a/docs/models/models.rst b/docs/models/models.rst
new file mode 100644 (file)
index 0000000..8cc6055
--- /dev/null
+++ b/docs/models/models.rst
@@ -0,0 +1,67 @@
+.. This work is licensed under a Creative Commons Attribution 4.0 International License.
+.. http://creativecommons.org/licenses/by/4.0
+.. (c) Anuket, The Linux Foundation, BIT Mesra, VTU and Others.
+
+
+==============================
+AI/ML Models for NFV Usecases.
+==============================
+
+This document describes all the models created by Anuket-Thoth project.
+
+*********************************
+1. Failure prediction (FP) models
+*********************************
+
+a. Summary of the VM Failure-Prediction models.
+===============================================
+We have developed Neural Network models for predicting failures in
+virtual machines (VMs) used in network function virtualization (NFV)
+environments by analysing VNF data. The data used to build these models
+are provided by Orange Labs, and the VMs are based on project Clearwater.
+
+The links for the data are:
+
+* Processed Data: https://drive.google.com/drive/folders/1crrVZMJwf00MP5qM7nmVEqFsatOAShla
+* Raw: https://drive.google.com/file/d/1QVipyoWPD1_4W_QXWzxEla4b88EWo5X5/view?usp=drivesdk
+* Raw Source: https://www.kaggle.com/datasets/imenbenyahia/clearwatervnf-virtual-ip-multimedia-ip-system
+
+
+These models are found under *models* directory. In the below table, only the jupyter-notebooks reference is given, which can be found in *models/failure_prediction/jnotebooks* folder . The corresponding python file can be found in *models/failure_prediction/python* folder.
+
+.. list-table:: Summary of Failure Prediction Models.
+   :widths: 25 25 25 100
+   :header-rows: 1
+
+   * - Model Name
+     - Failure type
+     - Source-File
+     - Comments
+   * - Decision Tree
+     - Virtual Machine
+     - Decision_Tree.ipynb
+     - Simplest (implementation-wise) case.
+   * - CNN
+     - Virtual Machine
+     - CNN.ipynb
+     - Convolutional Neural Network. Poorest among the Neural-network based models.
+   * - LSTM
+     - Virtual Machine
+     - LSTM.ipynb
+     - Basic Long Short-Term Memory. Better than CNN.
+   * - Attention LSTM
+     - Virtual Machine
+     - LSTM_attention.ipynb
+     - The attention mechanism distributes weights accordingly. Performance is similar to correlation LSTM.
+   * - Correlation LSTM
+     - Virtual Machine
+     - LSTM_correlation.ipynb
+     - Performs the best.
+   * - Correlation with Stacked LSTM
+     - Virtual Machine
+     - stacked_LSTM_correlation.ipynb
+     - Second best performance.
+   * - Correlation with Bi-LSTM
+     - Virtual Machine
+     - Bi_LSTMstacked_LSTM_Correlation.ipynb
+     - Third best performance.