Classification and Enrichment of Unlabeled Feedback Data using Machine Learning
B. Kranthi Kiran1, Padmaja Pulicherla2
1B. Kranthi Kiran, Associate Professor, JNTUHCEH. Kukatpally, Hyderabad, India.
2Padmaja Pulicherla , Professor, Department of CSE, Teegala Krishna Reddy Engineering College, Hyderabad, India.
Manuscript received on September 23, 2019. | Revised Manuscript received on October 15, 2019. | Manuscript published on October 30, 2019. | PP: 6647-6650 | Volume-9 Issue-1, October 2019 | Retrieval Number: A1912109119/2019©BEIESP | DOI: 10.35940/ijeat.A1912.109119
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: These days’ data gathered is unstructured. It is becoming very hard to have labelled data gathered, due to the volume of the data being generated every second. It is almost impossible to train a model on the unstructured/unlabelled data. The unlabelled data will be divided into groups using the ML techniques and CNN/Deep learning/Machine Learning techniques will be trained using the grouped data generated. The model will be enhanced over time by the feedback given by the users and with addition of new data as well. Existing models can be trained over labelled data only. Without labelled data models cannot be used for prediction and reinforcement learning. In this approach though the data is unlabelled if a feature column is specified we will be able to train the model with the help of SME. This will be helpful in many areas of classification and prediction of the trends and patterns. Machine learning, Deep learning techniques (Supervised) will be used to implement the data. Tools used will be Python, PyTorch and TensorFlow. Input can be any data (Audio/Video/pictographic/text). Labelled data and a model file which could be used for further predictions, and which will be improved over feedback.
Keywords: Classification, Unstrctured data, Machine Learning, CNN and Prediction.