2024 DevOps Lifecycle: Share your expertise on CI/CD, deployment metrics, tech debt, and more for our Feb. Trend Report (+ enter a raffle!).
Kubernetes in the Enterprise: Join our Virtual Roundtable as we dive into Kubernetes over the past year, core usages, and emerging trends.
Zone Leader at DZone
Hyderabad, IN
Joined Sep 2016
About
Business Analytics and Data Science consultant. Get in touch with him on Twitter: @sibanjandas
Stats
Reputation: | 3921 |
Pageviews: | 824.3K |
Articles: | 20 |
Comments: | 15 |
Expertise
AI/ML
Articles
Refcards
Understanding Data Quality
Trend Reports
Enterprise AI
In recent years, artificial intelligence has become less of a buzzword and more of an adopted process across the enterprise. With that, there is a growing need to increase operational efficiency as customer demands arise. AI platforms have become increasingly more sophisticated, and there has become the need to establish guidelines and ownership. In DZone’s 2022 Enterprise AI Trend Report, we explore MLOps, explainability, and how to select the best AI platform for your business. We also share a tutorial on how to create a machine learning service using Spring Boot, and how to deploy AI with an event-driven platform. The goal of this Trend Report is to better inform the developer audience on practical tools and design paradigms, new technologies, and the overall operational impact of AI within the business. This is a technology space that's constantly shifting and evolving. As part of our December 2022 re-launch, we've added new articles pertaining to knowledge graphs, a solutions directory for popular AI tools, and more.
Comments
Feb 14, 2020 · Sibanjan Das
Apr 22, 2019 · Sibanjan Das
Hi Phil, Thank you for your comments and appreciation. Glad that you liked the piece. Predicting future anomalies is an excellent thought.
For time series where we don't have future data, we need first to forecast the future values. Now, when we do some forecasting or prediction, there is an element of error associated with each prediction, and if we do Anomaly detection over these predictions, it might provide incorrect results. Anomaly detection is nothing but finding those points in the data that deviated from the usual observations.
As an extension to this logic, we can create a supervised anomaly detection model, which would be trained on the state of events right before an anomaly was detected(crime happened) and can be scored on the current set of attributes to identify whether the future behavior is anomalous or not( chance of a crime happening). This can help prevent crime before they happen.
I hope this helps answer your question.
Pre-arresting murderers can be challenged in the court of law, giving them a warning that the cops know their intention might be sufficient for them to stop doing the crime and keep them puzzled for a few days. :)
Regards
Sibanjan
Jun 13, 2018 · Sibanjan Das
Hello Thomas, Thank you pointing out the missing parameter. Somehow it got missed during editing. It is corrected now.
Nov 06, 2017 · Sibanjan Das
Thank You!! Here is the link for the data file and code. https://github.com/sibanjan/h2o
Oct 21, 2017 · Sibanjan Das
Perfect!! Thank You
Oct 21, 2017 · Sibanjan Das
Thanks Avkash! Glad that you liked it.
Oct 03, 2017 · Sarah Davis
Good One Anuj!!
Sep 24, 2017 · Sibanjan Das
HI Kriti, Depends on the R package you are using for modeling. Generally, it does not and you have to do the required transformations manually. There are many methods such as one hot encoding to transform categorical variable before creating an ML model. However, now a days we have certain R packages such as H2O which does this for you. Hope this helps. Thanks.
Sep 24, 2017 · Sibanjan Das
Aaron, Glad that it helped you. Served the purpose of my writing. Thank you!!
Jan 28, 2017 · Linda Gimmeson
Hello Linda, Nice to see my article being referred by you and you liked the content too. To add, we missed noting down Deep Learning which will also play a key role in automating Data Science in these Big Data and IoT environment.
Dec 29, 2016 · Sibanjan Das
Thanks Mateusz!! Glad that you liked it. I also feel OAA to be very powerful, scalable and handy than other tools.
Dec 23, 2016 · Sibanjan Das
Hello Deukjin!! Glad that you liked the article. Sorry to learn that the link to the "Excellent Presentatio" is not working. I could see that the link is prefixing a "dzone content" url before the slideshare link. Thanks for lettig us know that and we will fix it soon. For now, here is the link: http://www.slideshare.net/sbaltagi/why-apache-flink-is-the-4g-of-big-data-analytics-frameworks
Dec 22, 2016 · Sibanjan Das
:) . Thanks, Kellet. I'm glad you liked it. I moved down in the contest. But I am content thinking that my post would have inspired someone to write for the contest. haha..
Dec 08, 2016 · Sibanjan Das
Hello Karim, Thanks that you liked it. Sorry, I missed uploading the data and code. Will do that and post the link.
Dec 04, 2016 · Sibanjan Das
Hello Mr. Coulombe, Thanks for showing the typo. Also, for a valuable way to create a DN and the link. It helps!! Thank you once again.