https://github.com/amir22010/predicting-cloud-cpu-usage-on-azure-data
Forecasting future CPU Usage in Azure VM using Deep Learning Models. Compares LSTM , GRU and IndRNN
https://github.com/amir22010/predicting-cloud-cpu-usage-on-azure-data
Science Score: 10.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
○codemeta.json file
-
○.zenodo.json file
-
○DOI references
-
✓Academic publication links
Links to: arxiv.org -
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (7.3%) to scientific vocabulary
Last synced: 10 months ago
·
JSON representation
Repository
Forecasting future CPU Usage in Azure VM using Deep Learning Models. Compares LSTM , GRU and IndRNN
Basic Info
Statistics
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
- Releases: 0
Fork of amcs1729/Predicting-cloud-CPU-usage-on-Azure-data
Created over 5 years ago
· Last pushed almost 6 years ago
https://github.com/Amir22010/Predicting-cloud-CPU-usage-on-Azure-data/blob/master/
# Predicting Microsoft Azure cloud VM CPU Usage ## Project Name Predicting Microsoft Azure cloud VM CPU Usage ## Objective Predicting the CPU usage of cloud VM using data from the past. ## Why this might be worth considering? Well, Cloud services are becoming increasingly popular day by day. Server racks can be split between multiple users, with each user having anisolated sandbox (a virtual machine) for their application.A user can even have an application running across multiple VMs on separate machines. **Proactively allocating VMs can increase usage efficiency of underlying resources. If a computing cluster predicts the future resource usage of a user service will increase, it can preemptively scale up to accommodate a higher load. If it predicts that usage will decrease, it can deallocate VMs and save computing resources.** ## Dataset The Dataset contains the CPU Usage data of Microsoft Azure, which is a cloud service, sampled every 5 minutes. The data has three attributes - * Max CPU Utilization * Average CPU Utilization * Minimum CPU Utilization A glimpse into the data -  ## Models Used Three different types of models were trained and their results were compared. * LSTM Model ( 2 layers of 512 units each and a dense layer at the top) * GRU Model ( 2 layers of 512 units each and a dense layer at the top) * Independent RNN Model ( 2 layers of 512 units each and a dense layer at the top) ## Results ### LSTM MODEL ##### CPU High  ##### CPU Average  ##### CPU Low  ### GRU Model ##### CPU High  ##### CPU Avergae .png) ##### CPU Low .png) ### Independent RNN Model ##### CPU High  ##### CPU Average .png) ##### CPU Low .png) ## Model Performance The Model was evaluated using rolling window forecasting method. The Root Mean Squared Error (RMSE) , Mean Absolute Error (MAE) and the Mean Average Percentage Error (MAPE) were chosen as evaluation metrices. ##### LSTM Model * Test Score: 23987.87 RMSE * Test Score: 123.022565 MAE * Test Score: 1.004987 MAPE ##### GRU Model * Test Score: 24265.39 RMSE * Test Score: 125.725532 MAE * Test Score: 1.071244 MAPE ##### Independent RNN Model * Test Score: 30128.65 RMSE * Test Score: 131.286369 MAE * Test Score: 1.120023 MAPE ## Inference Based on the test scores, it is evident that the LSTM Model was better than the other two models. The error rate is only 1.004% which is quiet good and acceptable ## References [1] Independently Recurrent Neural Network (IndRNN): Building A Longer and Deeper RNN (https://arxiv.org/abs/1803.04831) [2] Keras implementation of IndRNN (https://github.com/titu1994/Keras-IndRNN)
Owner
- Name: Amir Khan
- Login: Amir22010
- Kind: user
- Location: India
- Repositories: 3
- Profile: https://github.com/Amir22010
working on developing a state of art AI solutions mainly in computer vision, chat bots and nlp domain. building an awesome AI as a professional developer 😍.