Overall Goal: Create a model to identify the customers more likely to accept offers.
1) Identify and justify statistically what is the best approach for our problem (classification -i.e. will accept or not-, regression – i.e. the ones that will leave more $-, or expected value -i.e. a x b). At the end our objective is the revenue (we prefer one customer using our product for $1000 than 10 customers using it for $80 for example)
2) Identify and justify the selection of the best model as well as the selection of the best features to be used (our dataset contains +300 features, we expect to end up with 5 to 8)
The work must be presented in a python notebook, with comments and proper justification (ideally statistically)
We have 3 previous campaigns results where we used a custom model. Using the output of this model will possibly bias the new model, for that we have executed a control group in our full customers ‘universe for which we also have the results. The data scientist should define a strategy indicating best way to maximize all the results we have.
42 pekerja bebas membida secara purata $2242 untuk pekerjaan ini
Hi Most of the previous work is very similar to your requirements. Python, ML and DL are my top skills. I'd like to be your long term partner. I can start right now. Thank you.