new dataset about dec-keras HOT 6 CLOSED

@dnnguifever You need to write a function to load your images to a numpy array with shape=[n, d], where n is the number of images and d is the dimension of each image. Please see datasets.py to understand the required format.

from dec-keras.

@XifengGuo, Can I test this algorithm on a non-image dataset? I have hourly temperature readings for 1 year (numpy array of shape 365 X 24), I need to cluster them, day-wise.

from dec-keras.

HiSamy, could you take each day's data to be an image? So, 24 pixels (hours) in each image (day), arranged in some kind of matrix, 6x4, say, or 24x1?

from dec-keras.

I want to make clusters of texts of different topics, which are unlabelled. Say I have notices about sport, economy and politics, would it be possible to train the DeepCluster for making this 3 clusters without any labelled text? what would i need, train first in reuters dataset?
Thanks for your work!

from dec-keras.

I have trained the reuters dataset with DEC.py and runned the experiment with run_exp.py, but I don't know how I can introduce my non-labelled dataset to cluster it. Could you help me with this please?

from dec-keras.

@basque21 The DEC itself is an unsupervised algorithm. So you can directly train DEC on your unlabeled data, just remember to remove the codes that require y. I recommend you try k-means first in order to understand how to cluster unlabeled data.

from dec-keras.