All activity
Rajiv Sambasivan
left a comment
I work in data science and ML, I am building a knowledge management library for data science. https://github.com/rajivsam/KMDS and https://github.com/rajivsam/kmds... and here is the tl;dr video: https://www.youtube.com/watch?v=ckr8YQJxF9I
Anyone from Cybersecurity teams, ML engineers, Data Scientists on here that are building something?
Haralds Gabrāns Zukovs
Join the discussion
Rajiv Sambasivan
started a discussion
Full real business dataset example with KMDS
Hello,
I have a full implementation of modeling analytics and machine learning workflows with the knowledge management tool that I have developed to model loan charge offs. The dataset is available from the data.gov website. The analytics use case is illustrated here:
https://github.com/rajivsam/kmds_recipes/wiki/A-simple-KMDS-analytics-reporting-project
The machine learning use case is...
Rajiv Sambasivan
left a comment
There are two guys who have written extensively about adoption of AI for business use cases. One is Foster Provost, the other is Eric Siegel. You should check out thier recommendations and guidelines for picking and measuring AI project success. The TL;DR version is to start with the business outcome you want to change with your AI idea and then have concrete, clear and interpretable measures...
Which AI business tools/ products will you actually use? How will you make that decision?
Catherine Gao
Join the discussion
Rajiv Sambasivan
started a discussion
Tool for capturing observations, rationale and decisions in data science projects.
Hello,
I am new here. Just wanted to post about a simple python package that I developed for recording your findings and observations in your data analysis projects. The repo is at https://github.com/rajivsam/KMDS
Recipes to use this in your data science projects are in https://github.com/rajivsam/kmds_recipes/wiki
Thank you, and best wishes to all.
Rajiv Sambasivan
left a comment
A lot of data science work is incremental and experimental. Any reasonable data science project requires choices that are mode based on the characteristics of the use case, and, there are many of these that accumalate over time. When it is time to rebuild the model because the data has changed this is forgotten. KMDS fixes this problem. See the kmds wiki for recipes to use kmds:...
KMDS
Knowledge Management for Data Science
KMDS is a python package that helps data scientists and data analysts collaborate and document their analysis and model development projects.See the kmds wiki for recipes to use kmds: https://github.com/rajivsam/kmds_recipes/wiki
KMDS
Knowledge Management for Data Science