Various algorithms for optimization and learning in adaptive systems

This thesis discusses various methods for learning and optimization in adaptive systems. Overall, it emphasizes the relationship between optimization, learning, and adaptive systems; and it illustrates the influence of underlying hardware upon the construction of efficient algorithms for learning and optimization. Chapter 1 provides a summary and an overview.

Chapter 2 discusses a method for using feed-forward neural networks to filter the noise out of noise-corrupted signals. The networks use back-propagation learning, but they use it in a way that qualifies as unsupervised learning. The networks adapt based only on the raw input data-there are no external teachers providing information on correct operation during training. The chapter contains an analysis of the learning and develops a simple expression that, based only on the geometry of the network, predicts performance.

Chapter 3 explains a simple model of the piriform cortex, an area in the brain involved in the processing of olfactory information. The model was used to explore the possible effect of acetylcholine on learning and on odor classification. According to the model, the piriform cortex can classify odors better when acetylcholine is present during learning but not present during recall. This is interesting since it suggests that learning and recall might be separate neurochemical modes (corresponding to whether or not acetylcholine is present). When acetylcholine is turned off at all times, even during learning, the model exhibits behavior somewhat similar to Alzheimer's disease, a disease associated with the degeneration of cells that distribute acetylcholine.

Chapters 4, 5, and 6 discuss algorithms appropriate for adaptive systems implemented entirely in analog hardware. The algorithms inject noise into the systems and correlate the noise with the outputs of the systems. This allows them to estimate gradients and to implement noisy versions of gradient descent, without having to calculate gradients explicitly. The methods require only noise generators, adders, multipliers, integrators, and differentiators; and the number of devices needed scales linearly with the number of adjustable parameters in the adaptive systems. With the exception of one global signal, the algorithms require only local information exchange.

Improving reporting of uncertainties in sea level rise assessments

Sea level rise (SLR) assessments are commonly used to identify the extent that coastal populations are at risk to flooding. However, the data and assumptions used to develop these assessments contain numerous sources and types of uncertainty, which limit confidence in the accuracy of modeled results. This study illustrates how the intersection of uncertainty in digital elevation models (DEMs) and SLR lead to a wide range of modeled outcomes. SLR assessments are then reviewed to identify the extent that uncertainty is documented in peer-reviewed articles. The paper concludes by discussing priorities needed to further understand SLR impacts. (PDF contains 4 pages)

Integrating climate change impacts to improve understanding of coastal climate change: heavy rains, strong winds, and high seas in coastal Hawaii, Alaska and the Pacific Northwest

Coastal storms, and the strong winds, heavy rains, and high seas that accompany them pose a serious threat to the lives and livelihoods of the peoples of the Pacific basin, from the tropics to the high latitudes. To reduce their vulnerability to the economic, social, and environmental risks associated with these phenomena (and correspondingly enhance their resiliency), decision-makers in coastal communities require timely access to accurate information that affords them an opportunity to plan and respond accordingly. This includes information about the potential for coastal flooding, inundation and erosion at time scales ranging from hours to years, as well as the longterm climatological context of this information. The Pacific Storms Climatology Project (PSCP) was formed in 2006 with the intent of improving scientific understanding of patterns and trends of storm frequency and intensity - “storminess”- and related impacts of these extreme events. The project is currently developing a suite of integrated information products that can be used by emergency managers, mitigation planners, government agencies and decision-makers in key sectors, including: water and natural resource management, agriculture and fisheries, transportation and communication, and recreation and tourism. The PSCP is exploring how the climate-related processes that govern extreme storm events are expressed within and between three primary thematic areas: heavy rains, strong winds, and high seas. To address these thematic areas, PSCP has focused on developing analyses of historical climate records collected throughout the Pacific region, and the integration of these climatological analyses with near-real time observations to put recent weather and climate events into a longer-term perspective.(PDF contains 4 pages)

Impact of radio broadcast on fish production and conservation in Kainji Lake basin

The decline in fishermen catches of Lake kainji, which went down from 28,639.6t in 1970 to 6.048t in 1974 (Ita 1982) prompted the promulgation of fisheries edict banning beach seines and other obnoxious fishing gear by Kebbi and Niger State governments in 1997 as a measure to prevent depletion of fisheries resources in the lake. Radio as a medium of communication was adopted as the method to disseminate information in the fisheries edict to the fisherfolks in order to bring about positive changes. Radio Niger, Koro station, was involved in the broadcast. Fisherfolk listened to the broadcast 32 times a week. Frame survey data and radio impact assessment studies were used to assess the effect of the broadcast on the number of beach seines and other obnoxious fishing gears on the lake which were banned in the fisheries edicts. It was revealed that the number of beach decreased drastically from 582 in 1997 to 122 in 1999 and to only 17 in 2001

Notes for guidance on preparation of research reports, research dissertations and scientific papers

The notes for guidance on preparation of research, research dissertations and scientific papers are meant to assist researchers on how to achieve a successful presentation, writing a technical report or research paper. This comprises some simple rules that have helped many inexperienced writers to get started, and have also helped more experienced writers to get out of a hopeless tangle of observations and inferences. (PDF contains 65 pages)