Advances in Artificial Intelligence: 11th Mexican by Grigori Sidorov, Sabino Miranda-Jiménez, Francisco

By Grigori Sidorov, Sabino Miranda-Jiménez, Francisco Viveros-Jiménez, Alexander Gelbukh (auth.), Ildar Batyrshin, Miguel González Mendoza (eds.)

The two-volume set LNAI 7629 and LNAI 7630 constitutes the refereed complaints of the eleventh Mexican foreign convention on man made Intelligence, MICAI 2012, held in San Luis Potosí, Mexico, in October/November 2012. The eighty revised papers offered have been rigorously reviewed and chosen from 224 submissions. the 1st quantity comprises forty papers representing the present major subject matters of curiosity for the AI group and their purposes. The papers are equipped within the following topical sections: desktop studying and trend popularity; laptop imaginative and prescient and photo processing; robotics; wisdom illustration, reasoning, and scheduling; clinical purposes of synthetic intelligence.

Show description

Read or Download Advances in Artificial Intelligence: 11th Mexican International Conference on Artificial Intelligence, MICAI 2012, San Luis Potosí, Mexico, October 27 – November 4, 2012. Revised Selected Papers, Part I PDF

Best international books

Tests and Proofs: 6th International Conference, TAP 2012, Prague, Czech Republic, May 31 – June 1, 2012. Proceedings

This ebook constitutes the refereed complaints of the sixth foreign convention on try and Proofs, faucet 2012, held in Prague, Czech Republic, in May/June 2012, as a part of the instruments 2012 Federated meetings. The nine revised complete papers provided including 2 invited papers, four brief papers and one instructional have been conscientiously reviewed and chosen from 29 submissions.

Cooperative Information Agents II Learning, Mobility and Electronic Commerce for Information Discovery on the Internet: Second International Workshop, CIA'98 Paris, France, July 4–7, 1998 Proceedings

This ebook constitutes the refereed court cases of the second one foreign Workshop on Cooperative details platforms, CIA'98, held in cognition with brokers global in July 1998 in Paris. The ebook offers 9 invited contributions including 14 revised complete papers chosen from a complete of fifty four submissions.

Data Warehousing and Knowledge Discovery: 12th International Conference, DAWAK 2010, Bilbao, Spain, August/September 2010. Proceedings

Info warehousing and information discovery has been commonly authorized as a key te- nology for organisations and enterprises to enhance their skills in info research, selection help, and the automated extraction of data from facts. With the exponentially turning out to be quantity of data to be integrated within the decision-making approach, the information to be thought of turn into an increasing number of advanced in either constitution and semantics.

Automated Reasoning: 6th International Joint Conference, IJCAR 2012, Manchester, UK, June 26-29, 2012. Proceedings

This publication constitutes the refereed lawsuits of the sixth foreign Joint convention on computerized Reasoning, IJCAR 2012, held in Manchester, united kingdom, in June 2012. IJCAR 2012 is a merger of prime occasions in automatic reasoning, particularly CADE (International convention on computerized Deduction), FroCoS (International Symposium on Frontiers of mixing Systems), FTP (International Workshop on First-Order Theorem Proving), and TABLEAUX (International convention on automatic Reasoning with Analytic Tableaux and comparable Methods).

Extra resources for Advances in Artificial Intelligence: 11th Mexican International Conference on Artificial Intelligence, MICAI 2012, San Luis Potosí, Mexico, October 27 – November 4, 2012. Revised Selected Papers, Part I

Example text

Nonetheless, we observed a few patterns that may prove helpful. Our experiments have demonstrated that initializations based on hierarchical clustering tend to find initial centroids that minimise the amount of iterations K-Means takes to converge. We can observe this in our experiments with the Agglomerative Ward and Divisive Ward criterion, but due to poor scalability these may not be the best options when dealing with large datasets. It seems that iK-Means and the Hartigan and Wong algorithm are the most scalable of the algorithms we experimented with, but the latter seems unable to minimise the number of K-Means iterations.

SK }, its centroid is the 1 + (k − 1) ∗ [N/K]th entity. Hartigan and Wong state that, using this method, no cluster will be empty after the first assignment of entities to centroids. We have formalized the algorithm as follows: 1. Order all N entities in the dataset in respect to the gravity centre. 2. For each cluster in S, set its centroid equal to 1 + (k − 1) ∗ [N/K]th entity. Hierarchical Agglomerative Ward. Ward’s hierarchical method [7] allows the creation of a dendrogram describing the whole dataset.

If not all data is clustered, go to Step 1. 5. Remove any cluster whose cardinality is below a user-specified threshold. 6. Using the found centroids, apply K-Means. In our experiments, we were not interested in finding the quantity of clusters in the datasets, which is a complex problem on its own; hence when applying iK-Means, we set its threshold to zero and chose the K largest clusters, where K is the given number of clusters. Where two clusters had the same cardinality, preference was given to the one that was found first, as this was the the most anomalous.

Download PDF sample

Rated 4.82 of 5 – based on 4 votes