Support Needed

[yousefkr]

Good Morning,

I am impressed with the results you have mentioned in your paper.

However, the reaction app http://indycar.demo.2.s3-website-us-east-1.amazonaws.com/ I can't see the RPM signal and when I switch between the cars, I only see sometimes a straight line representing the rpm. Is there a way to view the throttle and rpm through the whole race with the anomalies detected?

Thank you in advance for your time and help.

Best regards,
Yousef.

[chathurawidanage]

Hi Yousef,

We have migrated to a new react client. http://indycar.demo.3.s3-website-us-east-1.amazonaws.com/

We have containerized some components of the application to make it easily reproducible. Please go through the content of below folder.

https://github.com/DSC-SPIDAL/IndyCar/tree/master/containerize

Reards,
Chathura

[chathurawidanage]

Below repository has the changes to the original HTM java implementation we have discussed in the paper.

https://github.com/chathurawidanage/htm.java

However, we won't be able to share the IndyCar dataset with you.

[yousefkr]

Hello Chathura, Thank you very much for this information. In the focus view in the new react client. Is it possible to see the faults on the throttle or on the rpm? Or it is just visible on the overview like a warning and a danger. If it is possible could you please give me the car numbers that will have faults to focus on them. My reason for that question as I need to see what the algorithm detect as an error to make sure that this algorithm will work with my group project topic in the University. Thanks in advance for your help. Best regards, Yousef.

…

On Mon, Aug 10, 2020 at 3:50 PM Chathura Widanage ***@***.***> wrote: Hi Yousef, We have migrated to a new react client. http://indycar.demo.3.s3-website-us-east-1.amazonaws.com/ We have containerized some components of the application to make it easily reproducible. Please go through the content of below folder. https://github.com/DSC-SPIDAL/IndyCar/tree/master/containerize Reards, Chathura — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#16 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQRYMD5WJ257DPKCU2ZYNG3R773LTANCNFSM4PZMQFUQ> .

[chathurawidanage]

Hi,

We use HTM as the anomaly detection algorithm, and it outputs an anomaly score between 0 and 1, 1 being the most probable anomaly case. From the client application we categorize them to three color codes, green(score<0.3) yellow(score<0.6) and red(score>=0.6). If you need the raw values, you can observer them in realtime through the browser's web console by subscribing to the WebSocket stream.

Hope this will be helpful.

Regards,
Chathura

[zhangcheng]

I ran into this repo while searching ways to run Apache Storm on Kubernetes.
Would you please point me to the paper behind this project? @chathurawidanage @yousefkr
Thanks a lot.

[chathurawidanage]

Hey Zhang, Here you go. (13) (PDF) Anomaly Detection over Streaming Data: Indy500 Case Study (researchgate.net) <https://www.researchgate.net/publication/335499407_Anomaly_Detection_over_Streaming_Data_Indy500_Case_Study> Regards, Chathura

…

On Thu, Dec 16, 2021 at 1:25 PM ZHANG Cheng ***@***.***> wrote: I ran into this repo while searching ways to run Apache Storm on Kubernetes. Would you please point me to the paper behind this project? @chathurawidanage <https://github.com/chathurawidanage> @yousefkr <https://github.com/yousefkr> Thanks a lot. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#16 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABXZKCIEWNHT6WP42J7VXPDURGLOXANCNFSM4PZMQFUQ> .