Optical character recognition (also optical character reader, OCR) is the mechanical or electronic conversion of images of typed, handwritten or printed text into machine-encoded text, whether from a scanned document, a photo of a document, a scene-photo (for example the text on signs and billboards in a landscape photo) or from subtitle text superimposed on an image. (Source: wikipedia.com) Widely used as a form of information entry from printed paper data records whether passport documents, invoices, bank statements, computerised receipts, business cards, mail, printouts of static-data, or any suitable documentation it is a common method of digitising printed texts so that they can be electronically edited, searched, stored more compactly, displayed on-line, and used in machine processes such as cognitive computing, machine translation, (extracted) text-to-speech, key data and text mining. OCR is a field of research in pattern recognition, artificial intelligence and computer vision. (Source: wikipedia.com)

Data: The dataset contains 20,000 handwritten characters. The first column, of each row in the dataset file, is used to specify the class of the character (i.e. A, B, C etc.). The rest 16 columns are used for the corresponding feature vector which describes the specific character. The values in each feature vector are in range 0-15. All values are comma separated. Before you start working on this project, please read the bellow paper (p.161-165) to understand for what each feature stands for: P.W. Frey and D.J. Slate, Letter Recognition Using Holland style Adaptive Classifier, Machine Learning, Vol 6, 161-182 (1991). Example of data: T,2,1,5,6,13,9,2,5,11,10,4,5,2,1,7,9 I,5,12,15,3,6,8,5,9,9,4,12,2,5,7,4,9 The data file is uploaded in your Moodle Account.

Project Tasks:

Task 1 Give a quick description on how you will handle this problem with an MLP Neural Network. Explain: 1. How you will pre-process the data (input-output). 2. Why you can use Multilayer Perceptron (MLP) to solve this problem. 3. Which MLP parameter configurations do you need to choose for this problem? Why? 4. How will you evaluate your results?

Task 2(For this project you will use the KNIME platform.)Download the ocrData.txt file and pre-process the data (manually or through KNIME). During this procedure, you must normalize the input data and encode the output data. Describe how you normalize the input data and why. Describe how you encode the output data and why.

Task 3Use the KNIME Platform to solve this problem. You must develop a KNIME pipeline for MLP Neural Networks. In your report, you must present all KNIME configurations and comparative results for 3 different MLP Neural Network setups. You must present results related of Training Error, Testing Error and network accuracy. How the parameters you have chosen are affecting the results?