Вы находитесь на странице: 1из 22

Engineered for Tomorrow

A Project Presentation on
“Topic shadowing of News articles”

PRESENTED BY PROJECT GUIDE

Sayooj KS(1MJ15SCS737) Mr. Karthick Myilvahanan


Safdar Hashmi (1MJ15CS733) Asst prof, Dept of CSE

Srihari P (1MJ15CS743) MVJCE, Bangalore

5/8/2019
Engineered for Tomorrow

Contents
• Objectives.
• Literature survey.
• Diagrams.
• Dataset.
• Exploration and Feature Selection..
• Model Accurarcy.
• Connecting Database.
• Uploading Datasets.
• Validation of Datasets.

5/8/2019
Engineered for Tomorrow

Objectives
•To detect overshadowed news.
•To create overall statistical linkage between different news articles.
•To understand on average the type of news that reduces focus on important
news topics

5/8/2019
Engineered for Tomorrow

Literature Survey
Title “Prediction of kidney disease using data mining and machine learning technique”, Raj
prakash Thakur, Geeta Roy (IEEE)
Year Sept, 2017
Methodology 1. Machine Learning Techniques
2. K-means algorithm

Technology Machine Learning-Naïve Bayes, Maximum Entropy ,Enemble Classifier

Benefit The disease can be classified,but the accuracy of the system is low.

Issues It does not account for noise in the initial dataset

Future Work Adding feature to account for noise

5/8/2019
Engineered for Tomorrow

5/8/2019
Engineered for Tomorrow

Literature Survey
Title Classification of disease using data mining technique (IEEE) by Manoj Kumar Das

Year 2018
Methodology 1) Data mining technique
2) Classifying the data based on the trained model.

Technology Machine Learning


Data Mining
Existing system issues The existing system does not account for empty data in the dataset and hence the
accuracy an precession is low.
Benefit Entity selection and classification.

5/8/2019
Engineered for Tomorrow

5/8/2019
Engineered for Tomorrow

System Architecture

5/8/2019
Engineered for Tomorrow

Use Case Diagram

5/8/2019
Engineered for Tomorrow

Data Flow Diagram

Level 0 :

5/8/2019
Engineered for Tomorrow

Data Flow Diagram


Level 1 :

5/8/2019
Engineered for Tomorrow

Level 1

5/8/2019
Engineered for Tomorrow

5/8/2019
Engineered for Tomorrow

Sequence Diagram

5/8/2019
Engineered for Tomorrow

The Dataset
One of the most valuable assets a company has is data. The dataset is taken from
the UCI reprosatory that includes the dataset with 24 different features.
The dataset account for the various chronic kidney diseases as well as the non
chronic kidney disease.

The following sector is for defining a list which contains the required fields out
of the Raw DS (Kidney_disease.csv).

5/8/2019
Engineered for Tomorrow

5/8/2019
Engineered for Tomorrow

IMPLEMENTATION

5/8/2019
Engineered for Tomorrow

5/8/2019
Engineered for Tomorrow

5/8/2019
Engineered for Tomorrow

5/8/2019
Engineered for Tomorrow

5/8/2019
Engineered for Tomorrow

5/8/2019

Вам также может понравиться