Use a neural net or a convolutional neural net. Input data is 2000 real numbers each between -10 and 10. The output is 100 real number between -1 and 1. There are 30000 examples. Some need to be used as training data and some as test data in the development of the topology and weights of the network. I would like the response of the network to be within 1% of the training and test examples. I have some more test data to test the resulting neural net - do ensure that it has not been over-fitted to the dataset. The solution needs to include performance stats, final topology and weights, and also Matlab or R or (at a push) Python code of the network etc, so that I can rerun all code in order to validate all claims etc. I need somebody who is thorough and skillful.
45 freelancers are bidding on average £174 for this job
Hi... I am a Python and Machine Learning specialist, certified by Freelancer. I fully understand your project and I am sure I can help you. Let's discuss details by chat.
HI! I am interested in your project. If you give me this project, you will get a good result. thanks.... relevant skill; Machine Learning, Matlab and Mathematica, Python, Statistics
Hello We are highly motivated young technocrats from India. We are approachable, you can come to us if you have a struggle, we are happy to help. Team TiponTongue