Loading output library...
Loading output library...
Loading output library...
  • every row correspnds to a product, has 93 features, objective is to predict the target for the test data set*
  • the features are not explicitly decribed.
Loading output library...
Loading output library...
Loading output library...
Loading output library...
Loading output library...
Loading output library...
Loading output library...
Loading output library...
Loading output library...

Starting out with tensorflow

#Starting-out-with-tensorflow
Loading output library...
Loading output library...
Loading output library...
Loading output library...
Loading output library...
Loading output library...
Loading output library...

TO understand the dimensions

  • assuming you are doing by a batch of 100
  • dimensions:
  • x * W + b
  • 100,93 * 93,9 + ?,9
  • 100 * 9 + ?,9
  • numpy and python are smart enough to broadcast (replicate) bias to every row
  • in other words question mark is identified to be 100
Loading output library...
Loading output library...
Loading output library...
Loading output library...
Loading output library...