Derek C. Rose

A Fast and Stable Incremental Clustering Algorithm

Clustering is a pivotal building block in many data mining applications and in machine learning in general. Most clustering algorithms in the literature pertain to off-line (or batch) processing, in which the clustering process repeatedly sweeps through a set of data samples in an attempt to capture its underlying structure in a compact and efficient way. However, many recent applications require that the clustering algorithm be online, or incremental, in the that there is no a priori set of samples to process but rather samples are provided one iteration at a time. Accordingly, the clustering algorithm is expected to gradually improve its prototype (or centroid) constructs. Several problems emerge in this context, particularly relating to the stability of the process and its speed of convergence. In this paper, we present a fast and stable incremental clustering algorithm, which is computationally modest and imposes minimal memory requirements. Simulation results clearly demonstrate the advantages of the proposed framework in a variety of practical scenarios.

Deep Spatiotemporal Feature Learning with Application to Image Classification

Deep machine learning is an emerging framework for dealing with complex high-dimensionality data in a hierarchical fashion which draws some inspiration from biological sources. Despite the notable progress made in the field, there remains a need for an architecture that can represent temporal information with the same ease that spatial information is discovered. In this work, we present new results using a recently introduced deep learning architecture called Deep Spatio-Temporal Inference Network (DeSTIN). DeSTIN is a discriminative deep learning architecture that combines concepts from unsupervised learning for dynamic pattern representation together with Bayesian inference. In DeSTIN the spatiotemporal dependencies that exist within the observations are modeled inherently in an unguided manner. Each node models the inputs by means of clustering and simple dynamics modeling while it constructs a belief state over the distribution of sequences using Bayesian inference. We demonstrate that information from the different layers of this hierarchical system can be extracted and utilized for the purpose of pattern classification. Earlier simulation results indicated that the framework is highly promising, consequently in this work we expand DeSTIN to a popular problem, the MNIST data set of handwritten digits. The system as a preprocessor to a neural network achieves a recognition accuracy of 97.98% on this data set. We further show related experimental results pertaining to automatic cluster adaptation and termination.

A Public Key Cryptographic Method for Denial of Service Mitigation in Wireless Sensor Networks

The challenging characteristics of sensor nodes, including the constrained resources, the ad-hoc nature of their deployment and the vulnerability of wireless media, pose a need for unique security solutions. The advantages of Public Key Cryptography (PKC) for sensor network security are widely acknowledged and include resilience, scalability and decentralized management. Recent work has indicated that PKC is feasible in the wireless sensor network (WSN) environment, paving the way for many new security services and opportunities. However, the computational effort involved in performing PKC operations remains substantial. From an energy consumption perspective, it is imperative that the processing and communication resources be utilized only when required. To that end, PKC implementations are more vulnerable to Denial of Service (DoS) attacks, when compared to traditional security methods that require less resources. In particular, if a malicious party attacks a sensor node by repetitive requests to establish a key, the resources of the attacked node can be exhausted quite rapidly. In this paper, we propose a novel RSA-based framework for combating DoS attacks in WSN by ensuring that the malicious party will exhaust its resources prior to exhausting those of its counterparts. Under the proposed approach, the mathematical operations performed by the malicious party require two or three orders of magnitude more resources than those required by the attacked party. We also present three methodologies for establishing an ephemeral key, in which the proposed DoS mitigation mechanism is an embedded component. Implementation results on the Intel Mote 2 platform substantiate the clear advantages of the proposed method.

Self-certified public key generation on the intel mote 2 sensor network platform

It is widely acknowledged that security will play a key role in the successful design and deployment of wireless sensor networks (WSN). A prerequisite for achieving security is the ability to dynamically establish a secret key joint to two nodes. Elliptic Curve Cryptography (ECC) has emerged as a suitable public key cryptographic foundation for WSN. This poster will present an efficient ECC-based method for self-certified key generation in resource-constrained sensor nodes. In particular, we provide implementation results on the Intel Mote 2 sensor network platform which demonstrate that such key generation can be established in the order of 60 msec while consuming less than 30 mJ.

Applying deep-layered clustering to mammography image analytics

This paper details a methodology and preliminary results for applying a hierarchy of clustering units to mammographic image data. The identification of patients with breast cancer through the detection of microcalcifications and masses is a demanding classification problem; minimal false negatives are desired while simultaneously avoiding false positives that cause unnecessary cost to patients and health institutions. This research examines a segmented look at mammograms for computer aided detection with the goal of reliably labeling regions of interest requiring the attention of a radiologist. Classification is achieved by employing the building blocks, namely unsupervised clustering, of a deep learning architecture in tandem with a standard feed-forward neural network. Early results show promise for creating a classification engine that handles high-dimensional data with minimum engineering of image features, with a per-image patch sensitivity of 0.96 and specificity of 0.99.

Toward a Sequential Approach to Pipelined Image Recognition

This paper introduces a sequentially motivated approach to processing streams of images from datasets with low memory demands. We utilize fuzzy clustering as an incremental dictionary learning scheme and explain how the corresponding membership functions can be subsequently used in encoding features for image patches. We focus on replicating the codebook learning and classification stages from an established visual learning pipeline that has recently shown efficacy on the CIFAR-10 small image dataset. Experiments show that performance near batch oriented learning is achievable by combining naturally online learning mechanisms driven largely by stochastic gradient descent with strictly patch-wise operations. We further detail how back propagation can be used with a neural network classifier to modify parameters within the pipeline.

Multicast and quality of service provisioning in parallel shared memory switches

Growing demand for differentiated services and the proliferation of Internet multimedia applications requires not only faster switches/routers, but also the inclusion of guaranteed qualities of service (QoSs) and support for multicast traffic. Here, the authors introduce a parallel shared memory (PSM) architecture that addresses these demands by offering both QoS guarantees and support for multicast traffic. It is well known that PSM architectures represent an effective approach for distributing the high-memory bandwidth requirement found in output-queued (OQ) switches, while maintaining their desirable performance attributes. At the core of the PSM architecture is a memory management algorithm that determines, for each arriving packet, the memory unit in which it will be placed. The PSM architecture discussed should be considered with the context of fabric on a chip in mind, where an implementation is conceivable on a single chip, providing a plug-in emulated OQ switching solution. A description and detailed analysis of an efficient memory management algorithm that supports QoS and multicast traffic is given with a discussion of hardware implementation considerations that highlight the PSM architectures scalability and performance attributes.

7.2: Presentation session: Poster session and reception: “Applying deep-layered clustering to mammography image analytics”

This paper details a methodology and preliminary results for applying a hierarchy of clustering units to mammographic image data. The identification of patients with breast cancer through the detection of microcalcifications and masses is a demanding classification problem; minimal false negatives are desired while simultaneously avoiding false positives that cause unnecessary cost to patients and health institutions. This research examines a segmented look at mammograms for computer aided detection with the goal of reliably labeling regions of interest requiring the attention of a radiologist. Classification is achieved by employing the building blocks, namely unsupervised clustering, of a deep learning architecture in tandem with a standard feed-forward neural network. Early results show promise for creating a classification engine that handles high-dimensional data with minimum engineering of image features, with a per-image patch sensitivity of 0.96 and specificity of 0.99.