The deployed solution has these 2 steps: 1: The system trains on user provided historical incident data and builds & saves a deep learning model which is a representation of the historical data. 2: Once the model is generated, the solution can be used to predict next three most probable incidents. To achieve this end, the solution deploys the following 2 APIs over AWS Sagemaker:

1.Training API: The solution requires historical consecutive incidents along with contextual information as input for training the model. 2.Testing API: The solution requires recent historical incidents which would be used for forecasting next three incidents along with contextual information as input (same as training API).

Input

Supported content types: text/csv Sample input file: (https://tinyurl.com/y6nqt9u9 ) ** Following are the mandatory inputs for both the APIs:**

1. Incident Number: Unique ID for each incident
2. Incident Created Date: Date (MM/DD/YYYY) when Incident occurred.
3. Configuration: Category of Incident that occurred.

Note:

1. Two separate csv input files are required for training and testing.
2. For better results please provide at least 2000 consecutive historical incidents for training API.
3. Please provide at least 15 recent historical consecutive incidents for test API to forecast next three incidents.

Output

• Content type: text/csv • Sample output file:(https://tinyurl.com/y26dasjm ) ** The solution generates the following outputs:** Output contains three most probable forecasted incidents.

Invoking endpoint

AWS CLI Command

If you are using real time inferencing, please create the endpoint first and then use the following command to invoke it:

!aws sagemaker-runtime invoke-endpoint --endpoint-name $model_name --body fileb://$file_name --content-type 'text/csv' --region us-east-2 output.csv

Substitute the following parameters:

• "model-name" - name of the inference endpoint where the model is deployed
• file_name - input csv file name
• text/csv - content type of the given input
• output.csv - filename where the inference results are written to.

Resources

• Sample input
• Sample Output
• Sample Notebook

See Input Summary