<p>Artificial intelligence is a broad research area and there are many different reasons why it is interesting to study artificial intelligence. One of the main reasons is to understand how information might be represented in the human brain. The Recursive Auto Associative Memory (RAAM) is a connectionist architecture that with some success has been used for that purpose since it develops compact distributed representations for compositional structures.</p><p>A lot of extensions to the RAAM architecture have been developed through the years in order to improve the performance of RAAM; Bi coded RAAM (B-RAAM) is one of those extensions. In this work a modified B-RAAM architecture is tested and compared to RAAM regarding: Training speed, ability to learn with smaller internal representations and generalization ability. The internal representations of the two network models are also analyzed and compared. This dissertation also includes a discussion of some theoretical aspects of B-RAAM.</p><p>It is found here that the training speed for B-RAAM is considerably lower than RAAM, on the other hand, RAAM learns better with smaller internal representations and is better at generalize than B-RAAM. It is also shown that the extracted internal representation of RAAM reveals more structural information than it does for B-RAAM. This has been shown by hieratically cluster the internal representation and analyse the tree structure. In addition to this a discussion is added about the justifiability to label B-RAAM as an extension to RAAM.</p>

<p>As portable computers and wireless networks are becoming ubiquitous, it is natural to consider the user’s position as yet another aspect to take into account when providing services that are tailored to meet the needs of the consumers. Location aware systems could guide persons through buildings, to a particular bookshelf in a library or assist in a vast variety of other applications that can benefit from knowing the user’s position.</p><p>In indoor positioning systems, the most commonly used method for determining the location is to collect samples of the strength of the received signal from each base station that is audible at the client’s position and then pass the signal strength data on to a positioning server that has been previously fed with example signal strength data from a set of reference points where the position is known. From this set of reference points, the positioning server can interpolate the client’s current location by comparing the signal strength data it has collected with the signal strength data associated with every reference point.</p><p>Our work proposes the use of multiple successive received signal strength samples in order to capture periodic signal strength variations that are the result of effects such as multi-path propagation, reflections and other types of radio interference. We believe that, by capturing these variations, it is possible to more easily identify a particular point; this is due to the fact that the signal strength fluctuations should be rather constant at every position, since they are the result of for example reflections on the fixed surfaces of the building’s interior.</p><p>For the purpose of investigating our assumptions, we conducted measurements at a site at Växjö university, where we collected signal strength samples at known points. With the data collected, we performed two different experiments: one with a neural network and one where the k-nearest-neighbor method was used for position approximation. For each of the methods, we performed the same set of tests with single signal strength samples and with multiple successive signal strength samples, to evaluate their respective performances.</p><p>We concluded that the k-nearest-neighbor method does not seem to benefit from multiple successive signal strength samples, at least not in our setup, compared to when using single signal strength samples. However, the neural network performed about 17% better when multiple successive signal strength samples were used.</p>

Invariant object recognition is maybe the most basic and fundamental property of our visual system. It is the basis of many other cognitive tasks, like motor actions and social interactions. Hence, the theoretical understanding and modeling of invariant object recognition is one of the central problems in computational neuroscience. Indeed, object recognition consists of two different tasks: classification and identification. The focus of this thesis is on object identification under the basic geometrical transformations shift, scaling, and rotation. The visual system can perform shift, size, and rotation invariant object identification. This thesis consists of two parts. In the first part, we present and investigate the VisNet model proposed by Rolls. The generalization problems of VisNet triggered our development of a new neural network model for invariant object identification. Starting point for an improved generalization behavior is the search for an operation that extracts images features that are invariant under shifts, rotations, and scalings. Extracting invariant features guarantees that an object seen once in a specific pose can be identified in any pose. We present and investigate our model in the second part of this thesis.

Over 300,000 individuals are injured every year by vehicle related accidents in the United Kingdom alone. Government and the vehicle manufacturers are not only bringing new legislation but are also investing in vehicle safety research to bring this figure down. A private self-driven car is an important factor in maintaining the independence and quality of life of the third age individuals. However, since older people brings deterioration of cognitive, physical and visual abilities, resulting in slower reaction times and lapses while driving. The third age individuals are involved in more vehicle related accidents than middle aged individuals. This scenario is corrected by the fact that the number of third age individuals is increasing, especially in developed countries. It is expected that the percentage of third age individuals in the United Kingdom will increase to 20% of the total population by 2010. Several safety systems have been developed by the automotive industry including intelligent airbags, Electronic Stability Control (ESC) and pre-tensioned seat belts, but nothing has been specifically developed for the third age related problems. This thesis proposes a driver posture identification system using low resolution infrared imaging. The use of a low resolution thermal imager provides a reliable non-contact based posture identification system at a relatively low cost and is shown to provide robust performance over a wide range of conditions. The low resolution also protects the privacy of the driver. In order to develop the proposed safety system an Artificial Intelligent Thermal Imaging algorithm (AITl) is created in MatLAB. Experimentation is conducted in real and simulated environment, with human subjects, to evaluate the results of the algorithm. The result shows that the safety system is able to identify eighteen different driving postures. The system also provides other valuable information about the driver such as driver physical built, fatigue, smoking, mobile phone usage, eye-height and trunk stability. It is clear that in incorporating this safety system in the overall automotive central strategy, better safety for third age individual can be achieved. This thesis provides various contributions to knowledge including a novel neural network design, a safety system using low resolution infrared imager and an algorithm that can identify driver posture.

Caruthers argues that the mindreading capacity and the introspective capacity are in fact one and the same capacity. This single capacity relies on the same sub-personal "interpretive" mechanism that takes sensory information as input and produces attitudes as output. I use neuroscience research to show that if the “interpretive mechanism” exists, and moreover that it operates in accordance to Caruthers’ description in mindreading tasks, (e.g. detecting external cues and paying attention to others’ behavior), then this operation would have to be handled or implemented at the neural level by the Task Oriented Neural Network. On the other hand, it is well known that self-referential thought, including introspective thought is handled by the Default Mode Network. This consequence is problematic for the view that self and other attitude attributions are done by the same mechanism. The same cognitive operation can not be implemented by two distinct neural networks that are in competition with one another. Moreover, the Default Mode neural network and the Task Oriented networks implement such different types of thinking that they oppose and interrupt one another’s functioning. If the only difference between the two networks were that one simply handles a larger quantity of information than the other, then they wouldn’t be in competition. It appears that there is indeed something special about the very nature of self-referential information such that it determines the type of operations involved in its processing. / text

Manual sleep scoring, executed by visual inspection of the EEG, is a very time consuming activity, with an inherent subjective decisional component. Automatic sleep scoring could ease the job of the technicians, because faster and more accurate. Frequency information characterizing the main brain rhythms, and consequently the sleep stages, needs to be extracted from the EEG data. The approach used in this study involves a wavelet filter bank for the EEG frequency features extraction. The wavelet packet analysis tool in MATLAB has been employed and the frequency information subsequently used for the automatic sleep scoring by means of an artificial neural network. Finally, the automatic sleep scoring has been employed for epoching the fMRI data, thus allowing for studying brain resting state networks during sleep. Three resting state networks have been inspected; the Default Mode Network, The Attentional Network and the Salience Network. The networks functional connectivity variations have been inspected in both healthy and narcoleptic subjects. Narcolepsy is a neurobiological disorder characterized by an excessive daytime sleepiness, whose aetiology may be linked to a loss of neurons in the hypothalamic region.

The time spent searching for a parking space increases air pollution, driver frustration, and safety problems impacting among other issues, traffic congestion and as consequence the environment. In the United States, parking represents a $20 billion industry (National Parking Association, 2005), and research shows that a car is parked on average 90 percent of the time. To alleviate this problem, more parking facilities should be built or intelligent models to better utilize current facilities should be explored. In this thesis, a general methodology is proposed to provide solutions to the parking problem. First, stated preference data is used to study drivers' choice/behavior. Parking choices are modeled as functions of arrival time, parking price, age, income and gender. The estimated values show that choice is relatively inelastic with respect to distance and more elastic with respect to price. The data is used to estimate the price elasticity that induces drivers to change their behavior. Second, neural networks are used to predict space availability using data provided by a parking facility. The model is compared with traditional forecasting models used in revenue management. Results show that neural networks are an effective tool to predict parking demand and perform better than traditional forecasting models. Third, the price elasticity that induces drivers to change their choice or behavior is determined. Finally, taking as an input the forecasting results obtained from the neural network and the price elasticity, parking spaces are optimally allocated at different price levels to optimize facility utilization and increase revenue. This research considers a parking facility network consisting of multiple parking lots with two, three and four fare classes and utilizes revenue management techniques as a mean to maximize revenue and to stimulate and diversify demand. The output indicates the number of parking spaces that should be made available for early booking to ensure full utilization of the parking lot, while at the same time attempting to secure as many full price parking spaces to ensure maximization of revenue.

Though adaptational effects are found throughout the visual system, the underlying mechanisms and benefits of this phenomenon are not yet known. In this work, the visual system is modeled as a Deep Belief Network, with a novel “post-training” paradigm (i.e. training the network further on certain stimuli) used to simulate adaptation in vivo. An optional sparse variant of the DBN is used to help bring about meaningful and biologically relevant receptive fields, and to examine the effects of sparsification on adaptation in their own right. While results are inconclusive, there is some evidence of an attractive bias effect in the adapting network, whereby the network’s representations are drawn closer to the adapting stimulus. As a similar attractive bias is documented in human perception as a result of adaptation, there is thus evidence that the statistical properties underlying the adapting DBN also have a role in the adapting visual system, including efficient coding and optimal information transfer given limited resources. These results are irrespective of sparsification. As adaptation has never been tested directly in a neural network, to the author’s knowledge, this work sets a precedent for future experiments. / text

I present results from quantitative evaluation of the capability of hosting and trapping CO₂ of a carbonate brine reservoir from Dickman Field, Kansas. The analysis includes estimation of some reservoir parameters such as porosity and permeability of this formation using pre-stack seismic inversion followed by simulating flow of injected CO₂ using a simple injection technique. Liner et at (2009) carried out a feasibility study to seismically monitor CO₂ sequestration at Dickman Field. Their approach is based on examining changes of seismic amplitudes at different production time intervals to show the effects of injected gas within the host formation. They employ Gassmann's fluid substitution model to calculate the required parameters for the seismic amplitude estimation. In contrast, I employ pre-stack seismic inversion to successfully estimate some important reservoir parameters (P- impedance, S- impedance and density), which can be related to the changes in subsurface rocks due to injected gas. These are then used to estimate reservoir porosity using multi-attribute analysis. The estimated porosity falls within a reported range of 8-25%, with an average of 19%. The permeability is obtained from porosity assuming a simple mathematical relationship between porosity and permeability and classifying the rocks into different classes by using Winland R35 rock classification method. I finally perform flow simulation for a simple injection technique that involves direct injection of CO₂ gas into the target formation within a small region of Dickman Field. The simulator takes into account three trapping mechanisms: residual trapping, solubility trapping and mineral trapping. The flow simulation predicts unnoticeable changes in porosity and permeability values of the target formation. The injected gas is predicted to migrate upward quickly, while it migrates slowly in lateral directions. A large amount of gas is concentrated around the injection well bore. Thus my flow simulation results suggest low trapping capability of the original target formation unless a more advanced injection technique is employed. My results suggest further that a formation below our original target reservoir, with high and continuously distributed porosity, is perhaps a better candidate for CO₂ storage. / text

The interactive wireless sensor detects many hazardous gases such as Hexane, Propane, Carbon monoxide and Hydrogen. These gases are highly toxic and used in different kinds of manufacturing industries, domestic purpose and so on. So, building a sensor that can detect this kind of gases can save the environment; prevent the potential for explosion, and endangering human life. In long term, interactive wireless sensor can also prevent the financial losses that might occur due to the hazardous incident that might occur due to these toxic gases. Hexane is a colorless, strong gas which inhaled in significant amounts by a person then he may suffer with hexane poisoning and suffocation. It also causes skin burns when exposed in high concentrations. Propane, carbon monoxide and hydrogen can easily freeze in room temperature, if in contact with eye, it could permanently damage eye or cause blindness. The advantage of this wireless sensor is the use of artificial olfactory system (electronic nose) that can be taught to detect these hazardous gases. This sensor has a unique molecular combination of analysts, impurities and background that corresponds to a gas leak. It consists of a chemiresistor, such as an array of conductometric sensors, and a mechanism analyzing the data in real time. A smell-print is composed of many molecules which reaches receptor in the human nose. When a specific receptor receives a molecule, it sends a signal to the brain where the smell is identified and associated with that particular molecule. Similar manner, albeit substituting sensors for the receptors, and transmitting the signal to a machine learning algorithm for processing, rather than to the brain. This wireless gas leak sensing consists of microchip Pic 32, integrated electronic nose, automated data analysis unit, power supply, and communications. The communication channel will use the ZigBee link, or the cellular links, or other specific frequency wireless link. The time-stamped and position-stamped sensor measurement data are transmitted to the central computer in predetermined periods of time. The data will be stored in the computer database for possible future analysis of the gas leak development process.