Abstract -- Synthetic Aperture Radar(SAR) is a radar system used to create images of an object such as landscape. Classification of the SAR image, provides the magnitude of colors that illustrates the various feature of the SAR image. Combined classifier is proposed in this work to classify the SAR image. Proposed system consists of three processes including preprocessing, feature extraction and classification. Preprocessing stage uses bilateral filter to remove the noise present in the image. Daubechies4 wavelet is used to extract the feature by obtaining the approximated image. Instead of single classifier, the proposed classification process is completed using combined classifier which, improves the classification accuracy. Proposed combined classifier is the combination of three classifiers that are, Euclidean distance classifier, support vector machine(SVM) and Random forest classifier. Thus, 97.83% accuracy is achieved in classification using proposed method.

Abstract -- Power electronics is about converting the electrical power from the available form to the required form using power electronic circuits. Power electronic converters are classified as DC to DC Converters, DC to AC inverters, AC to DC rectifiers and AC to AC converters. Thyristorised time ratio controllers were used in the AC systems to implement power and frequency control. Phase angle controllers, extinction angle controllers and the integral cycle controllers are used in the making of AC to AC converters. The matrix converter is also an AC to AC converter. The topological structure of the matrix converter is in the form of a typical matrix with power electronic switches occupying the positions of a matrix structure. Matrix converters can convert AC to AC with modifications in frequency, amplitude, phase and number of phases. Matrix converters are classified as single phase converters and three phase converters. Matrix converters can be used for power control in lamps, heaters and speed control in electrical motors A matrix converter can also be used for converting from single phases to multiple phases. In this thesis the modeling & simulation of a single-phase matrix converter (SPMC) as a frequency changer modulated by the Sinusoidal Pulse Width Modulation (SPWM) subjected to passive load condition is presented. The proposed system is an AC to AC converter with features to change the frequency of the AC output voltage and frequency and the model was implemented using MATLAB/SIMULINK. Power MOSFETs and diodes were used as switching and direction control devices. The simulations have proved the efficacy of the proposed idea.

Abstract -- Breast cancer is a disease in which cells in the breast grow out of control. Breast cancer remains one of the top two diseases that lead to thousands of death in women every year. Since it is a very treacherous disease it's important to diagnose BC properly at early stages. Machine Learning algorithms are used to analyze any data set to extract data-driven model, prediction rule from the datasets. Also Data mining algorithms play an important role in prediction of early-stage breast cancer. Machine Learning methods gives an effective ways to classify data. Especially in the medical field, where those methods are widely used in diagnosis and analysis for decision making. So the main objective of our project is to assess the correctness in classifying the datasets.

Abstract -- The main objective of this project is to detect the underground cable faults and abnormalities occurring in underground cables using Raspberry Pi. The basic idea behind the working of this project is ohms’s law. The current would vary depending upon the length of fault of the cable. While a fault occurs due to many reasons in the cable, at a time of removing or repairing process, there is difficulty in locating the exact location of the fault. These fault details are after sent to any access point through the internet and displayed the output on the LCD display using GSM module. Thus it saves a lot of time, money and allows to service underground cable lines faster.

Abstract -- Using Energy efficiently in smart homes saves money, enhances sustainability and reduces carbon footprint at large. Consequently, the need for smart energy management is on the rise for smart homes and for smart cities in general. Residential Energy Management System is introduced and that was developed through Arduino, wireless communication technology. An efficient method is proposed to monitor and control power consumption levels in the home. The home appliances are connected to Wi-Fi hubs in the room, and these Wi-Fi hubs communicate the power consumption levels and energy usage data. Based on the received data and users requirement, control commands are initiated, and the home appliances can be controlled automatically from the home server. Load sharing information will be displayed on the home server and webpage. If the power consumption levels exceed the user-determined threshold, the system will automatically switch off, and warning is given to the consumer with a buzzer.