At Freiburger Energieoptimierung UG (www.frenop.de), DataRobot accelerated the pace of our analytics applied to energy efficiency and savings. Our team is integrated of Physicians who have a great knowledge of physical behaviors but we don’t have the time and the resources to develop a platform where we can analyze data from energy applications. DataRobot offers us automated solutions that are contributing for faster energy data analysis and also new discoveries with Machine Learning. The interface is very easy to use. Review collected by and hosted on G2.com.
DataRobot showed us that we need to have more data in order to proceed with some routines. Now we are investing more in time in finding this data in the real world so we can put it into these machine learning algorithms. Review collected by and hosted on G2.com.
Some structures are very easy to proceed here. Specially for robotics applications. Review collected by and hosted on G2.com.
Nothing to complain at the moment. I am thankful to find good applications for the programed structures that the program offers. Review collected by and hosted on G2.com.
DataRobot is a dream-come-true to a data scientist when it comes to modeling. The platform absolutely brute-forces through popular and niche modeling techniques with your data, present you with a ton of metrics afterwards on (top and all) results, and even has deployment options after whatever analysis you do. DataRobot presents you with a good avenue of ways to connect to your data: through a URL path, a local upload (such as csv or xlsx files), or through a hdfs path, so it's pretty easy to get started! Review collected by and hosted on G2.com.
There are limitations on how much data you plow through in one project if you're not on the enterprise edition (and there are still limitations on there too). Another thing that might not be great is that it is overly aggressive on the amount of models it runs through by default. Of course you can limit to a certain degree on which models DataRobot's autopilot will be run your data through, but if you're new to the platform, it'll take some ramp-up time on how to configure the autopilot correctly. One last thing, for data sources, they should've made some connectors to RDBMS type databases, or at least a couple of the popular ones like mysql or postgres. Review collected by and hosted on G2.com.
Its very help full to write the algorithm related to machine learning specially developer who want to give certain amount of data to module and model will predict based upon the data give. Review collected by and hosted on G2.com.
Start time is high,when you develop a heavy model some time its hanging Review collected by and hosted on G2.com.
Previously doing machine learning needs a lot of coding work on cleaning the data and building model, right now DataRobot can help us reduce this working effort and focus on improving the model and explaining the results. Review collected by and hosted on G2.com.
There are so many features on the interface, it's hard to find the feature you need. And there are so many professional concepts without detail explanation, not so user-friendly. Review collected by and hosted on G2.com.
Ability to easily build machine learning models algorithms ranging from simplistic regressions to complex gradient boosted trees Review collected by and hosted on G2.com.
- Interaction to the UI using scripting language like Python and R can still be made better
- Terminology can be standardized to the terms used in standard Data Science world (use crossvalidation instead of backtest) Review collected by and hosted on G2.com.
The best part of using Data Robot is that I am able to test out different models and change features, and it saves time. It is fast and provide good visualization. Review collected by and hosted on G2.com.
The biggest disappointment is that I cannot export the underlying code. Not sure if its not supported, but I did not find a way to do it. Review collected by and hosted on G2.com.