This work introduces a set of scalable algorithms to identify patterns of human daily behaviors. ... more This work introduces a set of scalable algorithms to identify patterns of human daily behaviors. These patterns are extracted from multivariate temporal data that have been collected from smartphones. We have exploited sensors that are available on these devices, and have identified frequent behavioral patterns with a temporal granularity, which has been inspired by the way individuals segment time into events. These patterns are helpful to both end-users and third parties who provide services based on this information. We have demonstrated our approach on two real-world datasets and showed that our pattern identification algorithms are scalable. This scalability makes analysis on resource constrained and small devices such as smartwatches feasible. Traditional data analysis systems are usually operated in a remote system outside the device. This is largely due to the lack of scalability originating from software and hardware restrictions of mobile/wearable devices. By analyzing the data on the device, the user has the control over the data, i.e. privacy, and the network costs will also be removed.
We address the problem of similarity search in large time series data- bases. We introduce a nove... more We address the problem of similarity search in large time series data- bases. We introduce a novel-dimensionality reduction technique that supports an indexing algorithm that is more than an order of magnitude faster than the previous best known method. In addition to being much faster our approach has numerous other advantages. It is simple to understand and implement, allows more
Proceedings of the sixth ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '00, 2000
ABSTRACT There has been much recent interest in adapting data mining algorithms to time series da... more ABSTRACT There has been much recent interest in adapting data mining algorithms to time series databases. Most of these algorithms need to compare time series. Typically some variation of Euclidean distance is used. However, as we demonstrate in this paper, Euclidean ...
Proceedings of the fifth ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '99, 1999
Page 1. Detecting Change in Categorical Data: Mining Contrast Sets Stephen D. Bay and Michael J. ... more Page 1. Detecting Change in Categorical Data: Mining Contrast Sets Stephen D. Bay and Michael J. Pazzani Department of Information and Computer Science University of California, Irvine Irvine, CA 92697, USA {sbay,pazzani}@ics.uci.edu ...
Proceedings of the 2002 SIAM International Conference on Data Mining, 2002
Time series are a ubiquitous form of data occurring in virtually every scientific discipline and ... more Time series are a ubiquitous form of data occurring in virtually every scientific discipline and business application. There has been much recent work on adapting data mining algorithms to time series databases. For example, Das et al. attempt to show how association rules ...
Page 1. ID2-of -3: Constructive Induction of M -of -N Concepts for Discriminators in Decision Tre... more Page 1. ID2-of -3: Constructive Induction of M -of -N Concepts for Discriminators in Decision Trees Patrick M. Murphy Dept. of Info. & Computer Science University of California, Irvine, CA 92717 pmurphy@ics.uci.edu Michael J. Pazzani Dept. of Info. ...
. Naive Bayesian classifiers utilise a simple mathematical modelfor induction. While it is known ... more . Naive Bayesian classifiers utilise a simple mathematical modelfor induction. While it is known that the assumptions on which thismodel is based are frequently violated, the predictive accuracy obtainedin discriminate classification tasks is surprisingly competitive in comparisonto more complex induction techniques. Adjusted probability naiveBayesian induction adds a simple extension to the naive Bayesian classifier.A numeric weight is inferred for each
Proceedings of the third annual conference on Autonomous Agents - AGENTS '99, 1999
... A long-term goal based on this research is the development of an intelligent ... up ways to p... more ... A long-term goal based on this research is the development of an intelligent ... up ways to process, evaluate and recommend information, leading to personalized information access without ... This paper describes the design and features of an information agent that we currently use ...
International Conference on Machine Learning, 1991
We discuss the types of noise that may occur in relational learning systems and describe two appr... more We discuss the types of noise that may occur in relational learning systems and describe two approaches to addressing noise in a relational concept learning algorithm. We then evaluate each approach expximentally.
... The general principle is to allow users to rate documents re-turned by the retrieval system w... more ... The general principle is to allow users to rate documents re-turned by the retrieval system with respect to their ... Taking word fre-quencies into account, maximum likelihood estimates for p(wt|cj; θ) can be derived from training data: ∑∑ ∑ ... 10.9 Trends in Content-Based Filtering ...
... The web provides an audience and a platform for individuals to express themselves or ... fo-c... more ... The web provides an audience and a platform for individuals to express themselves or ... fo-cuses on a dynamic user interface for personalized web-based news access: according ... describes Slider, a user inter-face specifically designed to capture users' preferences implicitly, in ...
We discuss algorithms for learning and revising user profiles that can determine which World Wide... more We discuss algorithms for learning and revising user profiles that can determine which World Wide Web sites on a given topic would be interesting to a user. We describe the use of a naive Bayesian classifier for this task, and demonstrate that it can incrementally learn profiles from user feedback on the interestingness of Web sites. Furthermore, the Bayesian classifier
International Conference on Machine Learning, 1998
Predicting items a user would like on the basis of other users' ratings for these items has b... more Predicting items a user would like on the basis of other users' ratings for these items has become a well-established strategy adopted by many recommendation services on the Internet. Although this can be seen as a classification problem, algo- rithms proposed thus far do not draw on results from the ma- chine learning literature. We propose a representation for collaborative
This work introduces a set of scalable algorithms to identify patterns of human daily behaviors. ... more This work introduces a set of scalable algorithms to identify patterns of human daily behaviors. These patterns are extracted from multivariate temporal data that have been collected from smartphones. We have exploited sensors that are available on these devices, and have identified frequent behavioral patterns with a temporal granularity, which has been inspired by the way individuals segment time into events. These patterns are helpful to both end-users and third parties who provide services based on this information. We have demonstrated our approach on two real-world datasets and showed that our pattern identification algorithms are scalable. This scalability makes analysis on resource constrained and small devices such as smartwatches feasible. Traditional data analysis systems are usually operated in a remote system outside the device. This is largely due to the lack of scalability originating from software and hardware restrictions of mobile/wearable devices. By analyzing the data on the device, the user has the control over the data, i.e. privacy, and the network costs will also be removed.
We address the problem of similarity search in large time series data- bases. We introduce a nove... more We address the problem of similarity search in large time series data- bases. We introduce a novel-dimensionality reduction technique that supports an indexing algorithm that is more than an order of magnitude faster than the previous best known method. In addition to being much faster our approach has numerous other advantages. It is simple to understand and implement, allows more
Proceedings of the sixth ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '00, 2000
ABSTRACT There has been much recent interest in adapting data mining algorithms to time series da... more ABSTRACT There has been much recent interest in adapting data mining algorithms to time series databases. Most of these algorithms need to compare time series. Typically some variation of Euclidean distance is used. However, as we demonstrate in this paper, Euclidean ...
Proceedings of the fifth ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '99, 1999
Page 1. Detecting Change in Categorical Data: Mining Contrast Sets Stephen D. Bay and Michael J. ... more Page 1. Detecting Change in Categorical Data: Mining Contrast Sets Stephen D. Bay and Michael J. Pazzani Department of Information and Computer Science University of California, Irvine Irvine, CA 92697, USA {sbay,pazzani}@ics.uci.edu ...
Proceedings of the 2002 SIAM International Conference on Data Mining, 2002
Time series are a ubiquitous form of data occurring in virtually every scientific discipline and ... more Time series are a ubiquitous form of data occurring in virtually every scientific discipline and business application. There has been much recent work on adapting data mining algorithms to time series databases. For example, Das et al. attempt to show how association rules ...
Page 1. ID2-of -3: Constructive Induction of M -of -N Concepts for Discriminators in Decision Tre... more Page 1. ID2-of -3: Constructive Induction of M -of -N Concepts for Discriminators in Decision Trees Patrick M. Murphy Dept. of Info. & Computer Science University of California, Irvine, CA 92717 pmurphy@ics.uci.edu Michael J. Pazzani Dept. of Info. ...
. Naive Bayesian classifiers utilise a simple mathematical modelfor induction. While it is known ... more . Naive Bayesian classifiers utilise a simple mathematical modelfor induction. While it is known that the assumptions on which thismodel is based are frequently violated, the predictive accuracy obtainedin discriminate classification tasks is surprisingly competitive in comparisonto more complex induction techniques. Adjusted probability naiveBayesian induction adds a simple extension to the naive Bayesian classifier.A numeric weight is inferred for each
Proceedings of the third annual conference on Autonomous Agents - AGENTS '99, 1999
... A long-term goal based on this research is the development of an intelligent ... up ways to p... more ... A long-term goal based on this research is the development of an intelligent ... up ways to process, evaluate and recommend information, leading to personalized information access without ... This paper describes the design and features of an information agent that we currently use ...
International Conference on Machine Learning, 1991
We discuss the types of noise that may occur in relational learning systems and describe two appr... more We discuss the types of noise that may occur in relational learning systems and describe two approaches to addressing noise in a relational concept learning algorithm. We then evaluate each approach expximentally.
... The general principle is to allow users to rate documents re-turned by the retrieval system w... more ... The general principle is to allow users to rate documents re-turned by the retrieval system with respect to their ... Taking word fre-quencies into account, maximum likelihood estimates for p(wt|cj; θ) can be derived from training data: ∑∑ ∑ ... 10.9 Trends in Content-Based Filtering ...
... The web provides an audience and a platform for individuals to express themselves or ... fo-c... more ... The web provides an audience and a platform for individuals to express themselves or ... fo-cuses on a dynamic user interface for personalized web-based news access: according ... describes Slider, a user inter-face specifically designed to capture users' preferences implicitly, in ...
We discuss algorithms for learning and revising user profiles that can determine which World Wide... more We discuss algorithms for learning and revising user profiles that can determine which World Wide Web sites on a given topic would be interesting to a user. We describe the use of a naive Bayesian classifier for this task, and demonstrate that it can incrementally learn profiles from user feedback on the interestingness of Web sites. Furthermore, the Bayesian classifier
International Conference on Machine Learning, 1998
Predicting items a user would like on the basis of other users' ratings for these items has b... more Predicting items a user would like on the basis of other users' ratings for these items has become a well-established strategy adopted by many recommendation services on the Internet. Although this can be seen as a classification problem, algo- rithms proposed thus far do not draw on results from the ma- chine learning literature. We propose a representation for collaborative
Uploads
Papers by Michael Pazzani