The machine learning has many capabilities one of them is classification. Classification employed in many contexts like telling hotel reviews good / bad, or inferring the image consists of dog, cat etc. As the data increases there is a need to organize it, for that purpose classification can be helpful. Modern classification problems involve the prediction of multiple labels simultaneously associated with a single instance known as "Multi Label Classification". In multi-label classification, each of the input data samples belongs to one or more than one classes or labels. The traditional binary and multi-class classification problems are the subset of the multi-label classification problem. In this paper we are implementing the multi label classification using CNN framework with keras libraries. Classification can be applied to different domain such as text, audio etc. In this paper we are applying classification for an image dataset.
Abstract-There is continues capture of large streaming data vital for application such as intensive health care system, Sensor networks, Object tracking etc.,. Data reduction of these huge data stream is carried out by similarity join processing which tracks the abnormal contents in real time data. The identification of anomalies such as abnormalities in Electro Cardio Gram (ECG) of an heart patient, predicting future casualties in weather monitor monitoring system, and providing heuristics in object tracking has to be effectively carried out. To achieve this we propose Identification of Anomalies in Time Series Data using Similarity Join Processing (IATSJ) to identify the anomalies by using Alternate Multilevel Segment Mean (AMSM) technique which reduces the data dimension and applying similarity join processing on these reduced data using sliding window concept. Experimental results show that, the time and space efficiency of our approach in anomaly detection from the given time series is better than the existing methods.
Electrical appliances most commonly consist of two electrical devices, namely, electrical motors and transformers. Typically, electrical motors are normally used in all sort of industrial purposes. Failures of such motors results in serious problems, such as overheat, shut down and
even burnt, in their host systems. Thus, more attention have to be paid in detecting the outliers. In a similar way, to avoid the unexpected power reliability problems and system damages, the prediction of the failures in the transformers is expected to quantify the impacts. By predicting
the failures, the lifetime of the transformers increases and unnecessary accidents is avoided. Therefore, this paper presents the detection of the outliers in electrical motors and failures in transformers using supervised machine learning algorithms. Machine learning techniques such as Support
Vector Machine (SVM), Random Forest (RF) and regression techniques like Support Vector Regression (SVR), Polynomial Regression (PR) are used to analyze the use cases of different motor specifications. Evaluation and the efficiency of findings are proved by considering accuracy, precision,
F-measure, and recall for motors. Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Square Error (RMSE) and R-squared Error (R2) are considered as metrics for transformers. The proposed approach helps to identify the anomalies like vibration loss, copper loss and overheating
in the industrial motor and to determine the abnormal functioning of the transformer that in turn leads to ascertain the lifetime. The proposed system analyses the behaviour of the electrical machines using the energy meter data and reports the outliers to users. It also analyses the abnormalities
occurring in the transformer using the parameters involved in the degradation of the paper-oil insulation system and the voltage of operation as a whole leads to the predict the lifetime.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.