# Meeting Log

## Weekly Meeting: 11/4
### PPT Section
* **IMPORTANT**: Framing must be changed to **Cloud Resource Usage Health Check**
* Predict output: Red

1. Motivation
* Hook: Story
* Story must be related with the Youtube video

2. Basic Introduction (Benefit)
* Reducing Cost
* Save Time
* Security: Fraud Detection
* Smart Strategy

3. System Architecture and Demo
4. Future Work

## Set Up: 8/16 2022 10:30 - 11:30
Members: Jay, Benjamin, Jessica, Sebastian, Winnie
### Goal
1. Project Planning
2. Set Up Goal
3. Weekly Planning
### Result
1. Job Assignments:
* __Model and Framework Design__: Benjamin, Jessica, Winnie
* __Data Visualization__: Brian, Sebastian
2. Weekly Goal:
* MTS Data visualization with ground truth
* Create GitHub repository
### TODO
- [x] Set up Github Repository
- [x] Research
- [x] Tableau
- [x] Power BI
- [x] Pyplot
- [x] Meeting 8/17 2022 10AM to sync up

## Daily: 8/17 2022 10:00 - 10:30
Members: Benjamin, Jessica, Sebastian
### Goal
1. Sync up research
2. Choose one platform to visualize
### Result
1. Decide to use Python based data visualization tools
2. Explore Orange Application for Data Visualization

### TODO
- [x] Improve graph on matplotlib
- [x] Explore Orange utilities
- [x] Prepare PPT for Friday

## Weekly: 8/19 2022 10AM
Members: Jay, Benjamin, Brian, Jessica, Sebastian, Winnie
### Goal
1. Weekly Progress Check
2. Weekly Planning
### Result

### TODO
- [ ] Decide plotly or matplotlib **-> transfer to backlog**
- [x] Decide weekly goal

## Daily: 8/23 2022 10 AM
Members: Jay, Benjamin, Brian, Jessica, Sebastian, Winnie
### Result
### TODO

## Daily: 8/25 2022 10 AM
Members: Jay, Benjamin, Jessica, Sebastian
### Result
* Push visualization to backlog and focus on visualization for research purposes

### TODO
- [x] Define functions
- [x] ARIMA: Winnie
- [x] XGboost: Sebastian
- [x] iForest: Jessica
- [x] SVM: Benjamin

## Weekly: 8/26 2022 10 AM
Members: Jay, Benjamin, Brian, Jessica, Sebastian, Winnie

### Result
* Model Framework
* Preprocessing
* Combine the 8 csv from RD for spatial
* Make a range between anomaly to **1**
* Model
* Make function for each of the model such that it can give an output given a dataframe and hyperparameter
* The output for UTS would be the anomaly score and the output for MTS would be the anomaly label (ie. 0 or 1)
* Postprocessing
* Manifold/Wighting: Make sure that the result from the temporal model and the spatial model are weighted equally.
* Consider rule based or clustering

### TODO
- [x] Make PPT

## Weekly: 9/2 2022 13:30
Members: Jay, Benjamin, Brian, Jessica, Sebastian, Winnie

### Result
1) next week ,after code revie every week data evaluatiopn slide(partiose github data)
2) f1 score is "useless" for now on, ignore him
3) jay will do a 甘特图
4) we will pressent at the first week of 12月
5) 11 月will start to pack, make video ,sliding ,and other importance stuf

### TODO
- [ ] Add ARIMA and XGboost to model.py
- [ ] Test out the models with SMAP, MSL, and SMD [link](https://github.com/NetManAIOps/OmniAnomaly)
- [ ] Dataset Visualization
- [ ] Postprocessing function

## UTS: 9/5 2022 (21:00 - 22:30)
Members: Jessica, Sebastian, Winnie

### Result
* Cleared up misunderstanding about the parameters `SW` and `lags`
* Sync up I/O of UTS Models:
* Input: MTS dataframe $df_{n \times C}$
* Output: Matrix $A_{(SW-n+1)\times C}$

### TODO
- [x] Add ARIMA to model.py
- [ ] Add Xgboost to model.py
- [ ] Make changes to Xgboost I/O

## Postprocessing: 9/7 2022 (21:00 - 00:00)
Members: Benjamin, Brian, Jessica, Winnie

### Result
1. UMAP 進度報告
2. PCA 進度報告
3. Up scaling 進度報告
4. 產出9/8的PPT

### TODO
1. visualization
2. 準備下週PPT

## Weekly: 9/8 2022 (17:40 - 18:40)
Members: Jay, Benjamin, Brian, Jessica, Sebastian, Winnie

### Result

### TODO

###### tags: `Documentation`