Binary Neural Networks (BNNs) enable efficient deep learning by saving on storage and computational costs. However, as the size of neural networks continues to grow, meeting computational require- ments remains a challenge. In this work, we propose a new form of quantization to tile neural network layers with sequences of bits to achieve sub-bit compression of binary-weighted neural networks. The method learns binary vectors (i.e. tiles) to populate each layer of a model via aggregation and reshaping operations. During in- ference, the method reuses a single tile per layer to represent the full tensor. We employ the approach to both fully-connected and convolutional layers, which make up the breadth of space in most neural architectures. Empirically, the approach achieves near full- precision performance on a diverse range of architectures (CNNs, Transformers, MLPs) and tasks (classification, segmentation, and time series forecasting) with up to an 8x reduction in size compared to binary-weighted models. We provide two implementations for Tiled Bit Networks: 1) we deploy the model to a microcontroller to assess its feasibility in resource-constrained environments, and 2) a GPU-compatible inference kernel to facilitate the reuse of a single tile per layer in memory.

Learning from the collective knowledge of data dispersed across private sources can provide neural networks with enhanced generalization capabilities. Federated learning, a method for collaboratively training a machine learning model across remote clients, achieves this by combining client models via the orchestration of a central server. However, current approaches face two critical limitations: i) they struggle to converge when client domains are sufficiently different, and ii) current aggregation techniques produce an identical global model for each client. In this work, we address these issues by reformulating the typical federated learning setup: rather than learning a single global model, we learn N models each optimized for a common objective. To achieve this, we apply a weighted distance minimization to model parameters shared in a peer-to-peer topology. The resulting framework, Iterative Parameter Alignment, applies naturally to the cross-silo setting, and has the following properties: (i) a unique solution for each participant, with the option to globally converge each model in the federation, and (ii) an optional early-stopping mechanism to elicit fairness among peers in collaborative learning settings. These characteristics jointly provide a flexible new framework for iteratively learning from peer models trained on disparate datasets. We find that the technique achieves competitive results on a variety of data partitions compared to state-of-the-art approaches. Further, we show that the method is robust to divergent domains (i.e. disjoint classes across peers) where existing approaches struggle.

Compressed Neural Networks have the potential to enable deep learning across new applications and smaller computational environments. However, understanding the range of learning tasks in which such models can succeed is not well studied. In this work, we apply sparse and binary-weighted Transformers to multivariate time series problems, showing that the lightweight models achieve accuracy comparable to that of dense floating-point Transformers of the same structure. Our model achieves favorable results across three time series learning tasks: classification, anomaly detection, and single-step forecasting. Additionally, to reduce the computational complexity of the attention mechanism, we apply two modifications, which show little to no decline in model performance: 1) in the classification task, we apply a fixed mask to the query, key, and value activations, and 2) for forecasting and anomaly detection, which rely on predicting outputs at a single point in time, we propose an attention mask to allow computation only at the current time step. Together, each compression technique and attention modification substantially reduces the number of non-zero operations necessary in the Transformer. We measure the computational savings of our approach over a range of metrics including parameter count, bit size, and floating point operation (FLOPs) count, showing up to a 53× reduction in storage size and up to 10.5× reduction in FLOPs.

Modern network infrastructures are in a constant state of transformation, in large part due to the exponential growth of Internet of Things (IoT) devices. The unique properties of IoT-connected networks, such as heterogeneity and non-standardized protocol, have created critical security holes and network mismanagement. In this paper we propose a new measurement tool, Intrinsic Dimensionality (ID), to aid in analyzing and classifying network traffic. A proxy for dataset complexity, ID can be used to understand the network as a whole, aiding in tasks such as network management and provisioning. We use ID to evaluate several modern network datasets empirically. Showing that, for network and device-level data, generated using IoT methodologies, the ID of the data fits into a low dimensional representation. Additionally we explore network data complexity at the sample level using Local Intrinsic Dimensionality (LID) and propose a novel unsupervised intrusion detection technique, the Weighted Hamming LID Estimator. We show that the algortihm performs better on IoT network datasets than the Autoencoder, KNN, and Isolation Forests. Finally, we propose the use of synthetic data as an additional tool for both network data measurement as well as intrusion detection. Synthetically generated data can aid in building a more robust network dataset, while also helping in downstream tasks such as machine learning based intrusion detection models. We explore the effects of synthetic data on ID measurements, as well as its role in intrusion detection systems.

The Internet of Things (IoT) is revolutionizing society by connecting people and devices seamlessly and providing enhanced user experience and functionalities. However, the unique properties of IoT networks, such as heterogeneity and non-standardized protocol, have created critical security holes and network mismanagement. We propose a new measurement tool for IoT network data to aid in analyzing and classifying such network traffic. We use evidence from both security and machine learning research, which suggests that the complexity of a dataset can be used as a metric to determine the trustworthiness of data. We test the complexity of IoT networks using Intrinsic Dimensionality (ID), a theoretical complexity measurement based on the observation that a few variables can often describe high dimensional datasets. We use ID to evaluate four modern IoT network datasets empirically, showing that, for network and device-level data generated using IoT methodologies, the ID of the data fits into a low dimensional representation; this makes such data amenable to the use of machine learning algorithms for anomaly detection.