Video Transcript:
Hello I'm Brandi and ammunition analytics and what combinations powerful visual tools. if you need a visual idea and hypothesis testing for a time like data specifically. for equity markets. because we have the largest litigation database. we are the only ones capable of utilizing the solution looking at data points. once we
have confirmed statistical significance we can use the tools we visually confirm our findings.
Let's run the simulation. for multivariate regression analysis we previously found a correlation between the average duration of closed cases and stock appreciation with about a five month gap from that life incorporated. the application is separated into two parts visualizers to help you visually search for and identify trends and machine learning tool to confirm such high pop. all data is loaded to CSV files never own pasta generate data files for any files that RJ index is supported data graph we simply choose the desired file from the drop-down. dimension which the graph and the time frame of the data to cut off.
Since we're looking for a the specific correlation you can grab the secondary active close duration in this case. inverse the data and set the month offset 25. this will let us visually examine the series in conjunction and determine if there is actually a the correlation here not in order to quantify this we're going to move onto the machine learning hypothesis tested. it works in a similar fashion. download the file and it's going to give us some options to choose our independent variables or predictors and our dependent variable or outcome. in this case we're using closer action to predict appreciation. at the time frame Markoff that 25. and we can use this drop-down select our machine learning bottle or pretty much the technique that the computer is going to use to use predictions. for those familiar with machine learning paradigm we've added the ability to treat the parameters and play around with the degrees of freedom that you are familiar.
With my creating a machine learning model promotes uses we
bought added the ability to automatically generate models. and let
the computer choose that's parameters for you. the last thing we have to worry about here is the iteration for both automatically and manually generative models the iterations will give us statistically significant numbers since. it's going to iterate randomly litigator and tested. for automatically generated models is going to determine the amount of parameters that have generated certificates so higher iterations will generally, lead to fine-tuning model of until seven points. once we have inputted all the parameters we simply click run model. let her run and we get the result inside model is automatically generated. the parameters is given to us as well as the percentage of accuracy that was evaluated fire bottle. there you have it with this tool were able to confirm and verify the correlations and we found previously from our multivariate regression analysis.
Premonition is an Artificial Intelligence system that mines Big Data to find out which lawyers, win which cases, in front of which judges.
It's creating a very, very unfair advantage in Litigation.
For more information, request a free demo: www.premonition.ai/take-a-tour/
For insider news on legal analytics and Premonition, subscribe: https://goo.gl/u4ZNdA
www.facebook.com/premonitionanalytics/
www.twitter.com/premonition_ai
www.linkedin.com/company/premonition-analytics