Breast cancer continues to be a significant health problem in the world. the most familiar breast anomalies types are mass and microcalcification. However Automatic methods for detecting these abnormalities can identi...
详细信息
ISBN:
(纸本)9781479959341
Breast cancer continues to be a significant health problem in the world. the most familiar breast anomalies types are mass and microcalcification. However Automatic methods for detecting these abnormalities can identify breast cancer at an early stage. In this paper, we propose a marker-controlled watershed algorithm to locate breast masses. the preprocessing step has been introduced to remove all undesirable areas from mammogram. Foreground and background markers are then selected in order to apply a watershed segmentation algorithm that identifies the location of tumor region in mammogram. the proposed method was successful to segment mass anomalies. It has been tested on publicly available Mammographic Image Analysis Society (MIAS) database and it has achieved an overall mass detection rate of 90.83% and an area Az of 0.913 under the receiver operating characteristic curve ROC for mass segmentation.
the proceedings contain 35 papers. the topics discussed include: hierarchical algorithm for pattern extraction from biological sequences;discoverying compounds that activate plant immunity using machine learning;a K-m...
ISBN:
(纸本)9781632665140
the proceedings contain 35 papers. the topics discussed include: hierarchical algorithm for pattern extraction from biological sequences;discoverying compounds that activate plant immunity using machine learning;a K-means based approach for the design of rotamer libraries: preliminary results;semi-supervised classification of protein sequences using burrows wheeler transformation-based features;advanced statistical analysis of RNA-seq differential gene expression profiles of animal gene-environment interactions;potential relationship discovery model base on context and its application in lung cancer field;analyzing conservation patterns and its influence on identifying protein functional sites;new weighting method of protein sequences for protein classification;and predicting protein subcellular localization with reduced feature set.
the final target of our research is to develop a system for forecasting local concentrated heavy rain, such as guerrilla rainstorms, by using whole sky images taken on the ground. To construct this system, this paper ...
详细信息
ISBN:
(纸本)9781628411867
the final target of our research is to develop a system for forecasting local concentrated heavy rain, such as guerrilla rainstorms, by using whole sky images taken on the ground. To construct this system, this paper proposes the following recognition methods: cloud amount, movement of clouds, and rain clouds. the experimental results show that red/blue (R/B) values are efficient for measuring the cloud amount. However, using the gravity of images and the difference among time-sequenced images is insufficient to recognize the movement of clouds and does not correlate well withthe R/B values and rain.
the proceedings contain 81 papers. the topics discussed include: variable hierarchical dependencies in feature selection;bi-spectrum based-EMD applied to the non- stationary vibration signals for bearing faults diagno...
ISBN:
(纸本)9781479959341
the proceedings contain 81 papers. the topics discussed include: variable hierarchical dependencies in feature selection;bi-spectrum based-EMD applied to the non- stationary vibration signals for bearing faults diagnosis;wireless landmines tracking system based on GPS and GPRS;performance of curvelets, dual-tree complex wavelet and discrete wavelet transform in handwritten word classification;possibility theory for supervised classification of remotely sensed images: a study case in an urban area in Algeria;proposition of a classification system beta-LS-SVM and its application to medical data sets;topological and textural features for off-line signature verification based on artificial immune algorithm;supervised wavelet-network based fuzzy-logic classifier performance on the UCI databases;scintigraphic image segmentation based on grammatical inference and spiral matrix;and microcalcifications detection in mammograms based on ant colony optimization and Markov random field.
In their primary structure, biological data are represented as a sequence of characters. the aim of this work is to automatically classify protein sequences. To do this, it seemed appropriate to use a data mining proc...
详细信息
Classical process discovery approaches usually investigate logs generated by processes in order to mine and discovery corresponding process schemas. When the collaboration processes case is addressed, such approaches ...
详细信息
ISBN:
(纸本)9781479959341
Classical process discovery approaches usually investigate logs generated by processes in order to mine and discovery corresponding process schemas. When the collaboration processes case is addressed, such approaches turn to be poorly effective, due to the fact that: (i) logs of collaboration processes are usually stored in heterogenous data storages which also expose different data types;(ii) it is not easy and direct to derive a common analysis model from such logs. As a consequence, classical methodologies usually fail. In order to fulfill this gap, in this paper we describe a composite methodology that combines semantics-based techniques and multidimensional analysis paradigms to support effective and efficient collaboration process discovery from log data.
the prime motivation for pattern discovery and machine learning research has been the collection and warehousing of large amounts of data, in many domains such as life sciences and industrial processes. Examples of un...
详细信息
Evolution of preventive safety devices for vehicles is highly expected to reduce the number of traffic accidents. Driver’s state adaptive driving support safety function may be one of solutions of the challenges to l...
详细信息
Withthe quick development of technology, the Internet of things is becoming one of hie most important components in our society. the Internet of things can make everything connected with Internet, and it is convenien...
详细信息
ISBN:
(纸本)9781479934348
Withthe quick development of technology, the Internet of things is becoming one of hie most important components in our society. the Internet of things can make everything connected with Internet, and it is convenient to recognize and manage In the process of recognizing the things in the Internet of things, it will appear irrelative and vague data;and we could not analyze and identity effectively. Focusing on the work of analyzing data, we propose PCA algorithm based technology in this paper, and apply it into works in the Internet of things through the theory and experimental results', we can prove that the technology we propose can select main information from prolix data, and complete the recognition works
this volume features key contributions from the internationalconference on patternrecognition Applications and Methods, (ICPRAM 2012,) held in Vilamoura, Algarve, Portugal from February 6th-8th, *** conference provi...
ISBN:
(纸本)9781493900923
this volume features key contributions from the internationalconference on patternrecognition Applications and Methods, (ICPRAM 2012,) held in Vilamoura, Algarve, Portugal from February 6th-8th, *** conference provided a major point of collaboration between researchers, engineers and practitioners in the areas of patternrecognition, both from theoretical and applied perspectives, with a focus on mathematical methodologies. Contributions describe applications of patternrecognition techniques to real-world problems, interdisciplinary research, and experimental and theoretical studies which yield new insights that provide key advances in the *** book will be suitable for scientists and researchers in optimization, numerical methods, computer science, statistics andfor differential geometers and mathematical physicists.
暂无评论