Simple and accurate decision tree based on fuzzy stop criteria approach

Babavalian, Mohammad Reza; Moghadam, Amir Masoud Eftekhari

doi:10.1109/ifsc.2013.6675583

Cited by 1 publication

(1 citation statement)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We have used DecisionTreeClassifier [19,20] Finding the Predictions/Classes: After fitting inputs into Classifier [21] model, we obtain classes predicted by our classifier. This task of prediction is done using predict() function of DecisionTreeClassifier class, which is stated below:…”

Section: Creation Of Classifiermentioning

confidence: 99%

A Study on Argument-Based Analysis of Legal Model

Sil

Saha

Roy

2021

Advances in Intelligent Systems and Computing

View full text Add to dashboard Cite

Unbounded delay in the delivery of justice has resulted in large number of pending cases thereby affecting the well-being of beneficiary in a broader sense. Due to this delay, complainant, accused or witness may either become unfit for trail or hostile, etc. which may jeopardize the promptness of judicial system. However, its efficiency can be enhanced using Machine Learning algorithm thereby reducing the workload of legal professional so that they can engage more time to resolve those pending cases. This paper focuses on argument-based prediction system over legal data set. Precisely, we have applied Supervised Machine Learning technique over the cases related to 'Domestic Violence Against Women' and have proposed a model to predict guilt of an accused. We have approached in sequential manner, precisely: (i) Hard-copies of argument based legal documents are collected from various trial courts of West Bengal. (ii) Data set are generated manually based on specific parameters which act as deciding factors for prediction of accusation. (iii) Supervised Machine Learning algorithm are used to train those legal data set to predict our desired output. We have validated the performance and accuracy of our proposed legal prediction system using standard classifiers like Decision Tree, Naive-Bayes Classifier and K-Nearest Neighbor (KNN).

show abstract

Section: Creation Of Classifiermentioning

confidence: 99%