如何使用Keras创建从表格数据进行训练的神经网络？

Question

I'm relatively new to Python, so please forgive the ignorance. I have tabular data that looks like:

Type,Name,Age,Breed1,Breed2,Gender,Color1,Color2,Color3,MaturitySize,FurLength,Vaccinated,Dewormed,Sterilized,Health,Quantity,Fee,State,RescuerID,VideoAmt,Description,PetID,PhotoAmt,AdoptionSpeed
2,Nibble,3,299,0,1,1,7,0,1,1,2,2,2,1,1,100,41326,8480853f516546f6cf33aa88cd76c379,0,Nibble is a 3+ month old ball of cuteness. He is energetic and playful. I rescued a couple of cats a few months ago but could not get them neutered in time as the clinic was fully scheduled. The result was this little kitty. I do not have enough space and funds to care for more cats in my household. Looking for responsible people to take over Nibble's care.,86e1089a3,1.0,2
2,No Name Yet,1,265,0,1,1,2,0,2,2,3,3,3,1,1,0,41401,3082c7125d8fb66f7dd4bff4192c8b14,0,I just found it alone yesterday near my apartment. It was shaking so I had to bring it home to provide temporary care.,6296e909a,2.0,0
1,Brisco,1,307,0,1,2,7,0,2,2,1,1,2,1,1,0,41326,fa90fa5b1ee11c86938398b60abc32cb,0,"Their pregnant mother was dumped by her irresponsible owner at the roadside near some shops in Subang Jaya. Gave birth to them at the roadside. They are all healthy and adorable puppies. Already dewormed, vaccinated and ready to go to a home. No tying or caging for long hours as guard dogs. However, it is acceptable to cage or tie for precautionary purposes. Interested to adopt pls call me.",3422e4906,7.0,3
1,Miko,4,307,0,2,1,2,0,2,1,1,1,2,1,1,150,41401,9238e4f44c71a75282e62f7136c6b240,0,"Good guard dog, very alert, active, obedience waiting for her good master, plz call or sms for more details if you really get interested, thanks!!",5842f1ff5,8.0,2

I have a ton of training data and I want to create a neural network to predict the last value AdoptionSpeed .

Here's what I have using keras so far:

from keras.models import Sequential
from keras.layers import Dense
import numpy

dataset = numpy.loadtxt("data/train.csv", delimiter=",")

X = dataset[:,0:8]
Y = dataset[:,8]

But I'm getting an error:

ValueError: could not convert string to float: Type

What am I doing wrong?

Answer 1

The default dtype of numpy loadtxt is float . Instead use:

dataset = np.loadtxt("data/train.csv", delimiter=",", dtype=np.str)

Answer 2

If you look at the csv file that you are passing to np.loadtxt funnction, some of the columns are not having float type of data. Few columns are strings. So, while loading the csv file, default datatype for conversion is float for np.loadtxt function. This is the reason for the error. Best way to overcome is by using python readlines function for a file and iterate over all the lines.

Generally, neural network expects inputs to be in the form of numbers. For converting the string to float values, you can use something like word2vec or Tf-Idf and their embedding alternatives.

As you want predict the Adoptionspeed , you can treat this problem as a regression problem.