Survey on the status of Computational Design research and practice

This report presents the results from a survey of current practice in the use of design optimization conducted amongst UK companies. The survey was completed by the Design Optimization Group in the Department of Engineering at Cambridge University. The general aims of this research were to understand the current status of design optimization research and practice and to identify ways in which the use of design optimization methods and tools could be improved.

Occupational practice and problems of artisanal fisherfolk in Oyan dam area of Ogun State

Artisanal Fish Societies constitutes one of the poorest societies in the developing world. Attempts to harness the potentials of such societies have often failed due to the enormity of the problem of poverty. This study was conducted in four major fishing villages namely: Abule Titun, Apojola, Imala Odo and Ibaro in order to investigate the occupational practices and the problems of rural artisanal fisherfolks in Oyam's Dam, area of Ogun State. Eighty respondents were randomly selected among the artisanal fisher folks for interview using interview guide. The findings revealed that 43.8% of the fisherfolks are within active age range of 31-40 years while 30% are within 21-30 years range. Also 31% had no formal education indicating a relatively high level of illiteracy among the fisherfolks while majority of the respondents practice fishing activities using paddle and canoe. It was similarly discovered from the study that the most pressing problems of the fishfolks is the lack of basic social amenities like electricity, potable water, access roads, hospital and markets. It is therefore recommended that basic social infrastructures be provided for the artisanal fishing communities in order to improve their social welfare, standard of living and the capacity to have a sustainable fishing occupation in the interest of food security and poverty alleviation

Optimal Data Distributions in Machine Learning

In the first part of the thesis we explore three fundamental questions that arise naturally when we conceive a machine learning scenario where the training and test distributions can differ. Contrary to conventional wisdom, we show that in fact mismatched training and test distribution can yield better out-of-sample performance. This optimal performance can be obtained by training with the dual distribution. This optimal training distribution depends on the test distribution set by the problem, but not on the target function that we want to learn. We show how to obtain this distribution in both discrete and continuous input spaces, as well as how to approximate it in a practical scenario. Benefits of using this distribution are exemplified in both synthetic and real data sets.

In order to apply the dual distribution in the supervised learning scenario where the training data set is fixed, it is necessary to use weights to make the sample appear as if it came from the dual distribution. We explore the negative effect that weighting a sample can have. The theoretical decomposition of the use of weights regarding its effect on the out-of-sample error is easy to understand but not actionable in practice, as the quantities involved cannot be computed. Hence, we propose the Targeted Weighting algorithm that determines if, for a given set of weights, the out-of-sample performance will improve or not in a practical setting. This is necessary as the setting assumes there are no labeled points distributed according to the test distribution, only unlabeled samples.

Finally, we propose a new class of matching algorithms that can be used to match the training set to a desired distribution, such as the dual distribution (or the test distribution). These algorithms can be applied to very large datasets, and we show how they lead to improved performance in a large real dataset such as the Netflix dataset. Their computational complexity is the main reason for their advantage over previous algorithms proposed in the covariate shift literature.

In the second part of the thesis we apply Machine Learning to the problem of behavior recognition. We develop a specific behavior classifier to study fly aggression, and we develop a system that allows analyzing behavior in videos of animals, with minimal supervision. The system, which we call CUBA (Caltech Unsupervised Behavior Analysis), allows detecting movemes, actions, and stories from time series describing the position of animals in videos. The method summarizes the data, as well as it provides biologists with a mathematical tool to test new hypotheses. Other benefits of CUBA include finding classifiers for specific behaviors without the need for annotation, as well as providing means to discriminate groups of animals, for example, according to their genetic line.