SEARCH · Search NASA
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
In this paper, we present a mathematical foundation, including a convergence analysis, for cascading architecture neural network.
Future applications of artificial intelligence to Mission Control Centers are presented in the form of the viewgraphs. The following subject areas are covered: basic objectives of the NASA-wide AI program; inhouse research program; constraint-based scheduling; learning and performance improvement for scheduling; GEMPLAN multi-agent planner; planning, scheduling, and control; Bayesian learning; efficient learning algorithms; ICARUS (an integrated architecture for learning); design knowledge acquisition and retention; computer-integrated documentation; and some speculation on future applications.
A new algorithm, called the CDF-Inversion Algorithm, is described. Using it, one can efficiently learn a probability distribution over a finite set to a specified accuracy and confidence. The algorithm can be extended to learn joint distributions over a vector space. Some implementation results are described.
A novel technique has been developed for anomaly detection of rocket engine test stand (RETS) data. The objective was to develop a system that postprocesses a csv file containing the sensor readings and activities (time-series) from a rocket engine test, and detects any anomalies that might have occurred during the test. The output consists of the names of the sensors that show anomalous behavior, and the start and end time of each anomaly. In order to reduce the involvement of domain experts significantly, several data-driven approaches have been proposed where models are automatically acquired from the data, thus bypassing the cost and effort of building system models. Many supervised learning methods can efficiently learn operational and fault models, given large amounts of both nominal and fault data. However, for domains such as RETS data, the amount of anomalous data that is actually available is relatively small, making most supervised learning methods rather ineffective, and in general met with limited success in anomaly detection. The fundamental problem with existing approaches is that they assume that the data are iid, i.e., independent and identically distributed, which is violated in typical RETS data. None of these techniques naturally exploit the temporal information inherent in time series data from the sensor networks. There are correlations among the sensor readings, not only at the same time, but also across time. However, these approaches have not explicitly identified and exploited such correlations. Given these limitations of model-free methods, there has been renewed interest in model-based methods, specifically graphical methods that explicitly reason temporally. The Gaussian Mixture Model (GMM) in a Linear Dynamic System approach assumes that the multi-dimensional test data is a mixture of multi-variate Gaussians, and fits a given number of Gaussian clusters with the help of the wellknown Expectation Maximization (EM) algorithm. The parameters thus learned are used for calculating the joint distribution of the observations. However, this GMM assumption is essentially an approximation and signals the potential viability of non-parametric density estimators. This is the key idea underlying the new approach.
A new learning algorithm termed cascade error projection (CEP) is presented. CEP is an adaption of a constructive architecture from cascade correlation and the dynamical stepsize of A/D conversion from the cascade back propagation algorithm.
We describe a neural network learning algorithm that implements differential learning in a generalized backpropagation framework. The algorithm regulates model complexity during the learning procedure, generating the best low-complexity approximation to the Bayer-optimal classifier allowed by the training sample.
The promise of active learning is to reduce the number of labeled examples required by supervised machine learning algorithms. The largest potential benefits lie in entirely new domains, for which no labeled examples yet exist. Yet to date, most active learning studies are retroactive and demonstrate the benefits that could have been gained if active learning had been used. What are the barriers to true adoption and utilization of active learning? We focus on two: (1) the cold start or class discovery problem, in which active learning methods may struggle to make progress with zero labeled examples, and (2) the cost of having the classifier in the loop to select the next example to be labeled. We assess different active learning approaches in the context of these two barriers and conclude with recommendations for how to employ active learning in new domains. As an example, we report on the use of active learning on a large, novel data set of Mars surface images.
This article shows how rational analysis can be used to minimize learning cost for a general class of statistical learning problems. We discuss the factors that influence learning cost and show that the problem of efficient learning can be cast as a resource optimization problem. Solutions found in this way can be significantly more efficient than the best solutions that do not account for these factors. We introduce a heuristic learning algorithm that approximately solves this optimization problem and document its performance improvements on synthetic and real-world problems.
I define the efficiency of a supervised learning strategy in classical estimation-theoretic terms.
Researchers interested in studying discrimination learning in primates have typically utilized variations in the Wisconsin General Test Apparatus (WGTA). In the present experiment, a new testing apparatus for the study of primate learning is proposed. In the video-task paradigm, rhesus monkeys (Macaca mulatta) respond to computer-generated stimuli by manipulating a joystick. Using this apparatus, discrimination learning-set data for 2 monkeys were obtained. Performance on Trial 2 exceeded 80 percent within 200 discrimination learning problems. These data illustrate the utility of the video-task paradigm in comparative research. Additionally, the efficient learning and rich data that were characteristic of this study suggest several advantages of the present testing paradigm over traditional WGTA testing.
Improved capabilities of Earth monitoring satellites are enabling a wide range of studies on the environmental effects of climate change, often leveraging the recent advancements in machine learning. At the same time, the new capabilities, including higher spatial resolution and temporal frequency, are expanding the amount of data generated at exponential rates. Further, a large majority of archived datasets generated by scientific processing is never used. This motivates the development of an efficient machine learning system for end-to-end processing of multi-level satellite datasets, from level 1 top of atmosphere observations to user friendly environmental variables of interest. Using current generation geostationary satellites GOES-16/17 (NOAA/NASA), and Himawari-8/9 (JAXA), we present an interchangeable set of machine models to perform spectral adjustment among sensors, physical model emulation, LEO-GEO emulation, and optical flow in a high performance computing environment. We use these tools on the NASA Earth eXchange (NEX) to generate consistent virtual observations across sensors, perform atmospheric correction and cloud detection, and estimate surface reflectance, surface temperature and atmospheric winds. This approach aims to improve the robustness of remotely sensed data processing by learning from diverse sets of observations while enabling near real-time and on-demand capabilities.
Intelligent tutoring systems (ITS) have been in existence for over a decade. However, few controlled evaluation studies have been conducted comparing the effectiveness of these systems to more traditional instruction methods. Two main promises of ITSs are examined: (1) Engender more effective and efficient learning in relation to traditional formats; and (2) Reduce the range of learning outcome measures where a majority of individuals are elevated to high performance levels. Bloom (1984) has referred to these as the two sigma problem; to achieve two standard deviation improvements with tutoring over traditional instruction methods. Four ITSs are discussed in relation to the two promises. These tutors have undergone systematic, controlled evaluations: (1) The LISP tutor (Anderson Farrell and Sauers, 1984); (2) Smithtown (Shute and Glaser, in press); (3) Sherlock (Lesgold, Lajoie, Bunzo and Eggan, 1990); and (4) The Pascal ITS (Bonar, Cunningham, Beatty and Well, 1988). Results show that these four tutors do accelerate learning with no degradation in final outcome. Suggestions for improvements to the design and evaluation of ITSs are discussed.
In this study artificial neural networks and fuzzy logic are used to control the jumping behavior of a three-link uniped robot. The biped locomotion control problem is an increment of the uniped locomotion control. Study of legged locomotion dynamics indicates that a hierarchical controller is required to control the behavior of a legged robot. A structured control strategy is suggested which includes navigator, motion planner, biped coordinator and uniped controllers. A three-link uniped robot simulation is developed to be used as the plant. Neurocontrollers were trained both online and offline. In the case of on-line training, a reinforcement learning technique was used to train the neurocontroller to make the robot jump to a specified height. After several hundred iterations of training, the plant output achieved an accuracy of 7.4%. However, when jump distance and body angular momentum were also included in the control objectives, training time became impractically long. In the case of off-line training, a three-layered backpropagation (BP) network was first used with three inputs, three outputs and 15 to 40 hidden nodes. Pre-generated data were presented to the network with a learning rate as low as 0.003 in order to reach convergence. The low learning rate required for convergence resulted in a very slow training process which took weeks to learn 460 examples. After training, performance of the neurocontroller was rather poor. Consequently, the BP network was replaced by a Cerebeller Model Articulation Controller (CMAC) network. Subsequent experiments described in this document show that the CMAC network is more suitable to the solution of uniped locomotion control problems in terms of both learning efficiency and performance. A new approach is introduced in this report, viz., a self-organizing multiagent cerebeller model for fuzzy-neural control of uniped locomotion is suggested to improve training efficiency. This is currently being evaluated for a possible patent by NASA, Johnson Space Center. An alternative modular approach is also developed which uses separate controllers for each stage of the running stride. A self-organizing fuzzy-neural controller controls the height, distance and angular momentum of the stride. A CMAC-based controller controls the movement of the leg from the time the foot leaves the ground to the time of landing. Because the leg joints are controlled at each time step during flight, movement is smooth and obstacles can be avoided. Initial results indicate that this approach can yield fast, accurate results.
At Marshall Space Flight Center a new capability has been established to aid the advancement of microfluidics for space flight monitoring systems. Lab-On-a-Chip Application Development (LOCAD) team has created a program for advancing Technology Readiness Levels (TRL) of 1 & 2 to TRL 6 and 7, quickly and economically for Lab-On-a-Chip (LOC) applications. Scientists and engineers can utilize LOCAD's process to efficiently learn about microfluidics and determine if microfluidics is applicable to their needs. Once the applicability has been determined, LOCAD can then perform tests to develop the new fluidic protocols which are different from macro-scale chemical reaction protocols. With this information new micro-devices can be created such as the development of a microfluidic system to aid in the search for life, past and present, on Mars. Particular indicators in the Martian soil can contain the direct evidence of life. But to extract the information from the soil and present it to the proper detectors requires multiple fluidic/chemical operations. This is where LOCAD is providing its unique abilities.
At Marshall Space Flight Center we have established a capability to investigate the use of microfluidics for space flight. The Lab-On-a-Chip Application Development (LOCAD) team has created a program for advancing Technology Readiness Levels (TRL) of 1 and 2 to TRL 6 and 7, quickly and economically for Lab-On-a-Chip (LOC) applications. Scientists and engineers can utilize LOCAD'S process to efficiently learn about microfluidics and determine if microfluidics is applicable to their needs. Once the applicability has been determined, LOCAD can then perform tests to develop the new fluidic protocols which are different from macro-scale chemical reaction protocols. With this information new micro-fluidic devices can be created and tested. Currently, LOCAD is focused on using microfluidics for both Environmental Monitoring & Control, and Medical Systems. Eventually, handheld portable units utilizing LOC technology will perform rapid tests to determine water quality, and microbial contamination levels. Since LOC technology is drastically reduced in physical size, it thereby reduces power, weight, volume, and sample requirements, a big advantage considering the resource constraints associated with spaceflight. Another one of LOCAD's current activities is the development of a microfluidic system to aid in the search for life on Mars.
Near-field radio holography is a common method for measuring and aligning mirror surfaces for millimeter andsub-millimeter telescopes. In instruments with more than a single mirror, degeneracies arise in the holographymeasurement, requiring multiple measurements and new fitting methods. We present HoloSim-ML, a Pythoncode for beam simulation and analysis of radio holography data from complex optical systems. This code usesmachine learning to efficiently determine the position of hundreds of mirror adjusters on multiple mirrors with fewmicrometer accuracy. We apply this approach to the example of the Simons Observatory 6 m telescope.
This paper explores the how's and why's of the Spitzer Mission Operations System's (MOS) success, efficiency, and affordability in comparison to other observatory-class missions. MOS exploits today's flight, ground, and operations capabilities, embraces automation, and balances both risk and cost. With operational efficiency as the primary goal, MOS maintains a strong control process by translating lessons learned into efficiency improvements, thereby enabling the MOS processes, teams, and procedures to rapidly evolve from concept (through thorough validation) into in-flight implementation. Operational teaming, planning, and execution are designed to enable re-use. Mission changes, unforeseen events, and continuous improvement have often times forced us to learn to fly anew. Collaborative spacecraft operations and remote science and instrument teams have become well integrated, and worked together to improve and optimize each human, machine, and software-system element.
In this paper, we present a mathematical foundation, including a convergence analysis, for cascading architecture neural networks. From this, a mathematical foundation for the casade correlation learning algorithm can also be found. Furthermore, it becomes apparent that the cascade correlation scheme is a special case of an efficient hardware learning algorithm called Cascade Error Projection.