Efficient learning of decomposable models with a bounded clique size

The learning of probability distributions from data is a ubiquitous problem in the fields of Statistics and Artificial Intelligence. During the last decades several learning algorithms have been proposed to learn probability distributions based on decomposable models due to their advantageous theoretical properties. Some of these algorithms can be used to search for a maximum likelihood decomposable model with a given maximum clique size, k, which controls the complexity of the model. Unfortunately, the problem of learning a maximum likelihood decomposable model given a maximum clique size is NP-hard for k > 2. In this work, we propose a family of algorithms which approximates this problem with a computational complexity of O(k · n^2 log n) in the worst case, where n is the number of implied random variables. The structures of the decomposable models that solve the maximum likelihood problem are called maximal k-order decomposable graphs. Our proposals, called fractal trees, construct a sequence of maximal i-order decomposable graphs, for i = 2, ..., k, in k − 1 steps. At each step, the algorithms follow a divide-and-conquer strategy based on the particular features of this type of structures. Additionally, we propose a prune-and-graft procedure which transforms a maximal k-order decomposable graph into another one, increasing its likelihood. We have implemented two particular fractal tree algorithms called parallel fractal tree and sequential fractal tree. These algorithms can be considered a natural extension of Chow and Liu’s algorithm, from k = 2 to arbitrary values of k. Both algorithms have been compared against other efficient approaches in artificial and real domains, and they have shown a competitive behavior to deal with the maximum likelihood problem. Due to their low computational complexity they are especially recommended to deal with high dimensional domains.

Forage strategy of juvenile Lates niloticus (L.) in Lake Victoria, Kenya

Six KMFRI stations located in Nyanza Gulf of Lake Victoria (Kenya) were sampled in order to investigate the forage strategy of juvenile Lates niloticus. Thirty speciemens were collected using a bottom trawl at each station and sorted into three size classes 1-2 cm and 3-20 cm total length. Stomach contents were analysed and taxonomic keys used to identify zoplankton and other insects. Caridina nilotica was the dominant food item in both frequency of occurrence and numerical abundance. In fish examined from 1-2 cm T.L., cladocerans were prominent food items, while at 2-3 and 3-20 cm, C. nilotica was dominant

Simple Coalitional Strategy Profiles in Repeated Games

This paper is a version of the discussion paper titled "Simple coalitional strategy profiles"

Reforming state-level coastal management and development policies: strategic retreat as an innovative, proactive and equitable coastal environmental management strategy

Atlantic and Gulf Coast shorelines include some of the most unique and biologically rich ecosystems in the United States that provide immeasurable aesthetic, habitat and economic benefits. Natural coastal ecosystems, however, are under increasing threat from rampant and irresponsible growth and development. Once a boon to local economies, complex natural forces – enhanced by global climate change and sea level rise - are now considered hazards and eroding the very foundation upon which coastal development is based. For nearly a century, beach restoration and erosion control structures have been used to artificially stabilize shorelines in an effort to protect structures and infrastructure. Beach restoration, the import and emplacement of sand on an eroding beach, is expensive, unpredictable, inefficient and may result in long-term environmental impacts. The detrimental environmental impacts of erosion control structures such as sea walls, groins, bulkheads and revetments include sediment deficits, accelerated erosion and beach loss. These and other traditional responses to coastal erosion and storm impacts- along with archaic federal and state policies, subsidies and development incentives - are costly, encourage risky development, artificially increase property values of high-risk or environmentally sensitive properties, reduce the post-storm resilience of shorelines, damage coastal ecosystems and are becoming increasingly unsustainable. Although communities, coastal managers and property owners face increasingly complex and difficult challenges, there is an emerging public, social and political awareness that, without meaningful policy reforms, coastal ecosystems and economies are in jeopardy. Strategic retreat is a sustainable, interdisciplinary management strategy that supports the proactive, planned removal of vulnerable coastal development; reduces risk; increases shoreline resiliency and ensures long term protection of coastal systems. Public policies and management strategies that can overcome common economic misperceptions and promote the removal of vulnerable development will provide state and local policy makers and coastal managers with an effective management tool that concomitantly addresses the economic, environmental, legal and political issues along developed shorelines. (PDF contains 4 pages)