ABSTRACT
Ever since the development of automatic sequence computers it has been possible for the machine to modify its own instructions, and this ability is the greatest single faculty in the complex that tempts the term "giant brain." Friedberg demonstrated a new technique in the modification of instructions; he allowed the machine to make alterations at "random," and lent direction to the maneuver by monitoring the result. This technique is far from being a feasible way to program a computer, for it took several hundred thousand errors before the first successful trial, and this was for one of the simplest tasks he could imagine. A simple principle of probabilities shows that a task compounded of two tasks of this same complexity would take several hundred thousand times as long, perhaps a million computer hours. It is the object of this study to examine techniques for abbreviating this process.
- fr1 R. M. Friedberg, "A learning machine," pt. I, IBM J. Res. and Dev., vol. 2, p. 2; January, 1958.Google Scholar
- fr2 A. G. Oettinger, "Programming a digital computer to learn," Phil. Mag., vol. 43, pp. 1243-1263; December, 1952.Google ScholarCross Ref
Recommendations
New machine learning algorithm: random forest
ICICA'12: Proceedings of the Third international conference on Information Computing and ApplicationsThis Paper gives an introduction of Random Forest. Random Forest is a new Machine Learning Algorithm and a new combination Algorithm. Random Forest is a combination of a series of tree structure classifiers. Random Forest has many good characters. ...
Statistical Properties of Pseudo Random Sequences and Experiments with PHP and Debian OpenSSL
Computer Security - ESORICS 2014AbstractNIST SP800-22 (2010) proposed the state of the art statistical testing techniques for testing the quality of (pseudo) random generators. However, it is easy to construct natural functions that are considered as GOOD pseudorandom generators by the ...
Experiments with random projections for machine learning
KDD '03: Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data miningDimensionality reduction via Random Projections has attracted considerable attention in recent years. The approach has interesting theoretical underpinnings and offers computational advantages. In this paper we report a number of experiments to evaluate ...
Comments