Papers by Michael J Watts
Finding relevant data in noisy, complex ecological times series: a comparison of two feature selection methods
Because of increasing transport and trade there is a growing threat of marine invasive species be... more Because of increasing transport and trade there is a growing threat of marine invasive species being introduced into regions where they do not presently occur. So that the impacts of such species can be mitigated, it is important to predict how individuals, particularly passive dispersers are transported and dispersed in the ocean as well as in coastal regions so that new incursions of potential invasive species are rapidly detected and origins identified. Such predictions also support strategic monitoring, containment and/or eradication programs. To determine factors influencing a passive disperser, around coastal New Zealand, data from the genus Physalia (Cnidaria: Siphonophora) were used. Oceanographic data on wave height and wind direction and records of occurrences of Physalia on swimming beaches throughout the summer season were used to create models using artificial neural networks (ANNs) and Naive Bayesian Classifier (NBC). First, however, redundant and irrelevant data were removed using feature selection of a subset of variables. Two methods for feature selection were compared, one based on the multilayer perceptron and another based on an evolutionary algorithm. The models indicated that New Zealand appears to have two independent systems driven by currents and oceanographic variables that are responsible for the redistribution of Physalia from north of New Zealand and from the Tasman Sea to their subsequent presence in coastal waters. One system is centred in the east coast of northern New Zealand and the other involves a dynamic system that encompasses four other regions on both coasts of the country. Interestingly, the models confirm, molecular data obtained from Physalia in a previous study that identified a similar distribution of systems around New Zealand coastal waters. Additionally, this study demonstrates that the modelling methods used could generate valid hypotheses from noisy and complicated data in a system about which there is little previous knowledge.
IAES International Journal of Artificial Intelligence, Jun 1, 2024
Financial technology (FinTech) which is included in the development of digitalization in the fina... more Financial technology (FinTech) which is included in the development of digitalization in the financial sector in the industrial era 4.0. FinTech can make any transactions anywhere with the pillars of peer-to-peer (P2P) lending, merchants, and crowdfunding. In the P2P lending pillar, there are borrowers and lenders who are digitized in FinTech devices. FinTech in Indonesia is controlled by a state agency called the financial services authority or otoritas jasa keuangan (OJK). In the movement of P2P lending, there are borrowers and lenders who can be said to be investors where these activities are reported to the OJK. This data can be forecasted using a neural network approach such as evolving connectionist system (ECoS), which is a method capable of forecasting with learning that develops in the hidden layer. In this research article, we present results on forecasting borrowers with a mean absolute percentage error (MAPE) of 0.148% and forecasting lenders with an accuracy measurement with MAPE of 0.209% with a learning rate 1=0.6 and a learning rate 2=0.3. So, this forecasting model can be said as an optimization in FinTech activities on the behavior of borrowers and lenders.
A student laptop roll-out for international information technology students
Adequate computing resources are essential to the effective teaching of Information Technology. T... more Adequate computing resources are essential to the effective teaching of Information Technology. There are several complicating factors when these resources are provided in the context of computer laboratories. These include the reliability of machines, consistency of software environments, and adequacy of hardware and the cost in both financial and human resources. We addressed these problems by progressively phasing out desktop computers in laboratories in favour of issuing laptops to IT students. These laptops were of a consistent specification and had a standard software environment. Practical problems encountered with this approach included procuring appropriate numbers of laptops in a timely manner, challenges with technical support and monitoring of students during practical tests and exams. Procedural problems included security of the laptops, handling returns and meeting student expectations. Each of these problems was solved and we succeeded in creating an efficient, cost-e...
The purpose of this paper is to describe and demonstrate a new approach for a practical and flexi... more The purpose of this paper is to describe and demonstrate a new approach for a practical and flexible environment for implementing knowledge and data fusion applications. Data fusion is used today in many engineering and managerial applications to help resolve complex planning, control and optimisation problems. A time-series application case study is presented and discussed: a prototype robot arm control example, utilising a fuzzy neural network (FuNN) tool module for off-line learning and rule manipulation, and a new on-line evolving fuzzy neural network (EFuNN) adapting tool module, from this environment.
Computational Ecology and Software, 2011
Existing cluster-based methods for investigating insect species assemblages or profiles of a regi... more Existing cluster-based methods for investigating insect species assemblages or profiles of a region to indicate the risk of new insect pest invasion have a major limitation in that they assign the same species risk factors to each region in a cluster. Clearly regions assigned to the same cluster have different degrees of similarity with respect to their species profile or assemblage. This study addresses this concern by applying weighting factors to the cluster elements used to calculate regional risk factors, thereby producing region-specific risk factors. Using a database of the global distribution of crop insect pest species, we found that we were able to produce highly differentiated region-specific risk factors for insect pests. We did this by weighting cluster elements by their Euclidean distance from the target region. Using this approach meant that risk weightings were derived that were more realistic, as they were specific to the pest profile or species assemblage of each r...
Table S1 Climate variables (units). Temperature first month of summer (C) Temperature second mont... more Table S1 Climate variables (units). Temperature first month of summer (C) Temperature second month of summer (C) Temperature third month of summer (C) Temperature first month of autumn (C) Temperature second month of autumn (C) Temperature third month of autumn (C) Temperature first month of winter (C) Temperature second month of winter (C) Temperature third month of winter (C) Temperature first month of spring (C) Temperature second month of spring (C) Temperature third month of spring (C) Summer temperature (C) Winter temperature (C) Annual temperature (C) Rainfall first month of summer (mm) Rainfall second month of summer (mm) Rainfall third month of summer (mm) Rainfall first month of autumn (mm) Rainfall second month of autumn (mm) Rainfall third month of autumn (mm) Rainfall first month of winter (mm)
Simple evolving connectionist systems and experiments on isolated phoneme recognition
2000 IEEE Symposium on Combinations of Evolutionary Computation and Neural Networks. Proceedings of the First IEEE Symposium on Combinations of Evolutionary Computation and Neural Networks (Cat. No.00EX448)
Applications: Neural Networks
The 2006 IEEE International Joint Conference on Neural Network Proceedings
2006 Sixth International Conference on Hybrid Intelligent Systems (HIS'06), 2006
Prioritizing the risk of plant pests by clustering methods; self-organising maps, k-means and hierarchical clustering
NeoBiota, 2013
Adapted conservation measures are required to save the Iberian lynx in a changing climate
Nature Climate Change, 2013
Proceedings of the 2001 Congress on Evolutionary Computation (IEEE Cat. No.01TH8546), 2001
The paper presents a method based on evolution strategies that attempts to optimise the training ... more The paper presents a method based on evolution strategies that attempts to optimise the training parameters of a class of on-line, adaptive connectionistbased learning systems called evolving connectionist systems (ECoS). ECoS are systems that evolve their structure and functionality through on-line, adaptive learning from incoming data. The ECoS paradigm is combined here with the paradigm of evolutionary computation to attempt to solve a difficult task of on-line adaptive adjustment and optimisation of the parameter values of the evolving system. Although the method presented is unsuccessful, some useful information about the properties of the ECoS model is still derived from the work.
Evolutionary optimisation of evolving connectionist systems
Proceedings of the 2002 Congress on Evolutionary Computation. CEC'02 (Cat. No.02TH8600), 2002
2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence), 2008
The apparent increase in number and magnitude of jellyfish blooms in the worlds oceans has lead t... more The apparent increase in number and magnitude of jellyfish blooms in the worlds oceans has lead to concerns over potential disruption and harm to global fishery stocks. Because of the potential harm that jellyfish populations can cause and to avoid impact it would be helpful to model jellyfish populations so that species presence or absence can be predicted. Data on the presence or absence of jellyfish of the genus Physalia was modelled using Multi-Layer Perceptrons (MLP) based on oceanographic data. Results indicated that MLP are capable of predicting the presence or absence of Physalia in two regions in New Zealand and of identifying significant biological variables.
Using time lagged input data to improve prediction of stinging jellyfish occurrence at New Zealand beaches by multi-layer perceptrons
Advances in Neuro-Information Processing. 15th International Conference, ICONIP 2008. Revised Selected Papers, 2009
Managing the long-term persistence of a rare cockatoo under climate change
Journal of Applied Ecology, 2012
Novel coupling of individual-based epidemiological and demographic models predicts realistic dynamics of tuberculosis in alien buffalo
Journal of Applied Ecology, 2011
Journal of Animal Ecology, 2013
1. Population viability analysis (PVA) is widely used to assess the extinction risk of threatened... more 1. Population viability analysis (PVA) is widely used to assess the extinction risk of threatened species and to evaluate different management strategies. However, conventional PVA neglects important biotic interactions and therefore can fail to identify important threatening processes. 2. We designed a new PVA approach that includes species interactions explicitly by networking species models within a single 'metamodel'. We demonstrate the utility of PVA metamodels by employing them to reinterpret the extinction of the carnivorous, marsupial thylacine Thylacinus cynocephalus in Tasmania. In particular, we test the claim that well-documented impacts of European settlement cannot account for this extinction and that an unknown disease must have been an additional and necessary cause. 3. We first constructed a classical, single-species PVA model for thylacines, which was then extended by incorporation within a dynamic predator-herbivore-vegetation metamodel that accounted for the influence of Europeans on the thylacine's prey base. Given obvious parameter uncertainties, we explored both modelling approaches with rigorous sensitivity analyses. 4. Single-species PVA models were unable to recreate the thylacine's extinction unless a high human harvest, small starting population size or low maximum population growth rate was assumed, even if disease effects were included from 1906 to 1909. In contrast, we readily recreated the thylacine's demise using disease-free multi-species metamodels that simulated declines in native prey populations (particularly due to competition with introduced sheep). 5. Dynamic, multi-species metamodels provide a simple, flexible framework for studying current species declines and historical extinctions caused by complex, interacting factors.
Information Sciences, 1998
Uploads
Papers by Michael J Watts
The Evolving Connectionist System (ECoS) is a class of contructive neural networks that are similar in the way in which neurons are added to their structures, and in the way in which their connection weights are modified. The ECoS algorithm is intended to address the problems with constructive neural networks.
Several problems with ECoS have been identified. These problems are: the excessive complexity of the Evolving Fuzzy Neural Network (EFuNN), which is the seminal ECoS network; the lack of a testable formalisation of ECoS; the dependence on fuzzy logic elements embedded within the network for fuzzy rule extraction; and the lack of methods for optimising ECoS networks.
This seminar describes how I addressed each of these problems in the course of my PhD research.