I need weak supervision approach for automatic labeling for my dataset. The problem is, I have labeled dataset with this feature text, verb,prep,start,end, label. For example weve had to break it down into what I think is askable about happy(text),
break(verb),down(prep),(15),(28)(1). The label is 1 for positive or true, and 0 for negative or false. Next I have huge unlabeled text dataset. The weak supervision model (or any else) have to label this dataset for example it has to extract So we take a large collection of images and we break them down into their little image patches(text), break(verb),down(prep),(48),(63)(1) for positive and When organic material dies in nature microbes and bacteria break it and shut down into nutrient rich soil completing the life cycle(text) break(verb),down(prep),(48),(63)(0) for negative. The model should have accuracy, at least 50 %, noisy labeling still be accepting, as long it able to extract and labeling.
I am willing to pay more, if i see promising result. and please explain your approach.
4 freelanceri licitează în medie 96$ pentru acest proiect
Hello, I think the best way to deal with this project is merging natural language processing techniques with machine learning algos. NLP will help us in extracting good set of candidates and can guide ML algo in a bet Mai multe
iam an NLP expert, do kindly share the details so that we can connect over chat, i can help you close the task within 3 days if is to auto classify into a particular category or tag
Hi, I am an engineer whose expertise is on deep learning and computer vision. I have sufficient knowledge about NLP and very good knowledge about deep learning algorithms, etc. I have designed several unsupervised a Mai multe