ccpclassifier

Machine Learning approach using Python and Google's TensorFlow to classify CCP's.

Program Requirements:

• Must have TensorFlow installed (this was developed for the GPU version)
• A folder named "predictions" in the directory with the python scripts to write machine learning predictions to
• Python 3

Program Overview:

• masterscript.py
• Script to be run. Change variables under the section labeled...

___________________________________________________________________________

DATA PREPARATION

___________________________________________________________________________

numCells - the number of cell directories folderPath - the path to the folder with all of the cell folders labeled 'CellN_Ns' NOTE: if data is not found in folders labeled 'CellN_Ns', change the values in trackloader.py dropCols: Array of columns to not use. Acceptable names include: ['trackNum', 'frame', 'lifetime', 'max_intensity', 'background', 'totaldisp', 'max_msd', 'catIdx', 'aux', 'avg_rise', 'avg_dec', 'risevsdec', 'avg_mom_rise', 'avg_mom_dec', 'risevsdec_mom']