Deep neural networks (DNNs) underpin many machine learning applications. Production quality DNN models achieve high inference accuracy by training millions of DNN parameters which has a significant resource footprint. This presents a challenge for resources operating at the extreme edge of the network, such as mobile and embedded devices that have limited computational and memory resources. To address this, models are pruned to create lightweight, more suitable variants for these devices. Existing pruning methods are unable to provide similar quality models compared to their unpruned counterparts without significant time costs and overheads or are limited to offline use cases. Our work rapidly derives suitable model variants while maintaining the accuracy of the original model. The model variants can be swapped quickly when system and network conditions change to match workload demand. This paper presents DNNShifter, an end-to-end DNN training, spatial pruning, and model switching system that addresses the challenges mentioned above. At the heart of DNNShifter is a novel methodology that prunes sparse models using structured pruning. The pruned model variants generated by DNNShifter are smaller in size and thus faster than dense and sparse model predecessors, making them suitable for inference at the edge while retaining near similar accuracy as of the original dense model. DNNShifter generates a portfolio of model variants that can be swiftly interchanged depending on operational conditions. DNNShifter produces pruned model variants up to 93x faster than conventional training methods. Compared to sparse models, the pruned model variants are up to 5.14x smaller and have a 1.67x inference latency speedup, with no compromise to sparse model accuracy. In addition, DNNShifter has up to 11.9x lower overhead for switching models and up to 3.8x lower memory utilisation than existing approaches.

Collaborative machine learning (CML) techniques, such as federated learning, were proposed to collaboratively train deep learning models using multiple end-user devices and a server. CML techniques preserve the privacy of end-users as it does not require user data to be transferred to the server. Instead, local models are trained and shared with the server. However, the low resource utilisation of CML techniques makes the training process inefficient, thereby limiting the use of CML in the real world. Idling resources both on the server and devices due to sequential computation and communication is the principal cause of low resource utilisation. A novel framework PipeLearn that leverages pipeline parallelism for CML techniques is developed to improve resource utilisation substantially. A new training pipeline is designed to parallelise the computations on different hardware resources and communication on different bandwidth resources, thereby accelerating the training process in CML. The pipeline is further optimised to ensure maximum utilisation of available resources. The experimental results confirm the validity of the underlying approach of PipeLearn and highlight that when compared to federated learning: (i) the idle time of the server can be reduced by 2.2x - 28.5x, (ii) the network throughput can be increased by 56.6x - 321.3x, and (iii) the overall training time can be accelerated by 1.5x - 21.6x under varying network conditions for two popular convolutional models without sacrificing accuracy. PipeLearn is available for public download from https://github.com/blessonvar/PipeLearn.

Ensemble Decision Systems offer a unique form of decision making that allows a collection of algorithms to reason together about a problem. Each individual algorithm has its own inherent strengths and weaknesses, and often it is difficult to overcome the weaknesses while retaining the strengths. Instead of altering the properties of the algorithm, the Ensemble Decision System augments the performance with other algorithms that have complementing strengths. This work outlines different options for building an Ensemble Decision System as well as providing analysis on its performance compared to the individual components of the system with interesting results, showing an increase in the generality of the algorithms without significantly impeding performance.