Spam - classifier tensorflow
Budget: $30 – $250 USD
Dataset:
4.6k records, 1.8k spam
Columns, e.g. word_freq_make, word_freq_address, capital_run_length_average, etc.
Tasks:
1. build a fully connected classifier network that will classify the mails.
2. check which parameters affect decision the most (biggest gradient wrt to answer)
3. build a classifier that doesn't use 1, 2, 10 most significant parameters. Does it still perform better then random?
I would need a full python notebook with the 3 tasks above done for the attached dataset. The dataset is from the following link: https://archive.ics.uci.edu/ml/datasets/spambase
It needs to be done usin tensorflow and the output should be a python notebook. I have attached the start of the workbook you can continue
Dataset description:
Our collection of spam e-mails came from our postmaster and individuals who had filed spam. Our collection of non-spam e-mails came from filed work and personal e-mails, and hence the word 'george' and the area code '650' are indicators of non-spam. These are useful when constructing a personalized spam filter. One would either have to blind such non-spam indicators or get a very wide collection of non-spam to generate a general purpose spam filter.
4.6k records, 1.8k spam
Columns, e.g. word_freq_make, word_freq_address, capital_run_length_average, etc.
Tasks:
1. build a fully connected classifier network that will classify the mails.
2. check which parameters affect decision the most (biggest gradient wrt to answer)
3. build a classifier that doesn't use 1, 2, 10 most significant parameters. Does it still perform better then random?
I would need a full python notebook with the 3 tasks above done for the attached dataset. The dataset is from the following link: https://archive.ics.uci.edu/ml/datasets/spambase
It needs to be done usin tensorflow and the output should be a python notebook. I have attached the start of the workbook you can continue
Dataset description:
Our collection of spam e-mails came from our postmaster and individuals who had filed spam. Our collection of non-spam e-mails came from filed work and personal e-mails, and hence the word 'george' and the area code '650' are indicators of non-spam. These are useful when constructing a personalized spam filter. One would either have to blind such non-spam indicators or get a very wide collection of non-spam to generate a general purpose spam filter.