Algorithm Optimization - Reviewed
We are a small group of private individuals developing financial software.
We are looking for 3 or 4 experts in algorithms, mathematics, statistics and/or data science with a high level experience in the underlying technologies. If you are not an expert or do not have significant experience please do not bid.
We require multiple experts as we intend to leverage multiple solutions across a variety of methodologies/strategies.
We require you to design and build an algorithm (or a combination of methodologies and algorithms) that can be run in your selected technology to deliver a set of queries that will meet the objectives (detailed below).
Query = Set or restrictions to be applied to a set of data to generate a subsample
We expect each solution to be made up of 4-6 queries targeting different regions in the data.
This project is a proof of concept firstly utilizing 3 days data (in sample) then a further 3 days data (out of sample, please see the milestones) and will be fully awarded for completion to all selected bidders. Understanding your solution is vital, when applying the resulting queries out of sample (milestone 2) they must maintain a reasonable aggregate performance.
We expect to select 3 to 4 freelancers. Once this project is completed successfully we will need to work with each of you to scale up your solution to be applied to 30 days data. We may leverage cloud technology for this however this is not in scope for this project. Once this is completed we have a further 4 markets we need to repeat this for immediately.
• We will provide one datasets with one target variable (“Score”), a timestamp and 24 independent variables. The dataset contains ~55 thousand observations
• The goal is to generate a query (set of restrictions) on the independent variables. Each set of restrictions will return a subsample of the dataset on which we evaluate an objective function.
• Your restrictions can be greater than or less than, or be combined to define a range. At most only two restriction on each variable.
• Specifically, the objective function is the sum of the target of the observations in the selected subsample. Each query (set of restrictions) has to return at least 10 valid responses and the sum total must be a positive result.
• In addition, any observations that come less than 60 seconds after a valid observation in this subsample will be removed. So each query has to return at least 10 responses that are 60 seconds apart from each other.
• In other words, your goal in this project is cornering an estimated 6 regions of the dataset using intervals on the independent variables, and maximize the density of positive values of the target.
• We have an indicative result (this will not be made available) we are looking for successful solutions to exceed.
• Solutions must provide evidence of performance testing (Milestone 3). This can agreed but should be achieved by altering the restrictions by n% reflecting a degradation in performance but not zero performance
• You can find the dataset in the Excel file “[url removed, login to view]”.
• You can find example restrictions in the excel file “Example Restrictions”
FURTHER INFORMATION AND TIPS ATTACHED IN "ALGORITHM OPTIMIZATION - REVIEWED"
18 freelancers are bidding on average $612 for this job
I am an expert in Matlab and Math. I have many experiences in optimization. I am an expert in dynamic/sophisticated approaches. I can do it correctly. Regards.