Theses and Dissertations

Permanent URI for this collectionhttp://ir.daiict.ac.in/handle/123456789/1

Browse

Filters

2017 - 201912020 - 20211

Settings

Subject: search.filters.subject.Classification

Search Results

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    Credit Card Fraud Detection Using MachineLearning Algorithms
    (Dhirubhai Ambani Institute of Information and Communication Technology, 2021) Tank, Ekta; Das, Manik Lal
    Credit Card payment facilitating people to pay for goods and service quickly. The credit payment is gaining popularity day by day because of its benefits. With the popularity of credit card payment, crime related to credit card fraud is also increasing. Credit card fraud leads to a colossal amount of loss of financial institutions like banks and the customer. Detecting fraud is costly and time-confusing, Though it is too important to detect fraud and prevent fraud in the future. In this thesis, the challenges of credit card fraud are discussed. Credit Card fraud is treated as the classification problem, and experiments are carried out with Decision Tree, Random Forest and SVM. Credit Card data will always be highly imbalanced in nature, having fewer number of frauds than normal transactions. To deal with this problem, resampling techniques are performed on the dataset. The credit card fraud problem is also considered an anomaly detection problem having fraud as an anomaly. The main objective of the research is to find an effective approach to detect fraud. This thesis compares the classification approach with the anomaly detection approach. Also, classification results are tried to improve using data level resampling techniques. Comparison results are discussed in Result Section.
  • Thumbnail Image
    ItemOpen Access
    Human Action Recognition Using Deep Neural Networks
    (Dhirubhai Ambani Institute of Information and Communication Technology, 2017) Thakkar, Shaival; Joshi, Manjunath V.
    "In this thesis, we present a hierarchical approach for human action classification using 3-D Convolutional neural networks (3-D CNN). Human actions refer to positioning and movement of hands and legs and hence can be classified based on those performed by hands or by legs or, in some cases, both. This acts as the intuition for our work on hierarchical classification. In this work, we consider the actions as tasks performed by hand or leg movements. Therefore, instead of using a single 3-D CNN for classification of given actions, we use multiple networks to perform the classification hierarchically, that is, we first classify an action into a hand or leg action and then use two separate networks for hand and leg action classes to perform classification among target action categories. In particular, we train three networks to classify six different actions, comprising of three actions each for hands and legs. The use of 3-D CNN enables automatic extraction of features in spatial as well as temporal domain, avoiding the need for hand crafted features. This makes it one of the better approaches when it comes to video classification. We use the KTH dataset to evaluate our approach and comparison with the state of the art methods shows that our approach outperforms most of the state of the art methods."