Artificial Intelligence

Deterministic Degradation Process for Diffusion GAN and Its Inversion

Authors: Jeongik Cho
Comments: 10 Pages.

Recently, diffusion models have shown impressive generative performance. However, they have the disadvantage of having a high latent dimension and slow sampling speed. To increase the sampling speed of diffusion models, diffusion GANs have been proposed. But the latent dimension of diffusion GANs using non-deterministic degradation is still high, making it difficult to invert the generative model. In this paper, we introduce an invertible diffusion GAN that uses deterministic degradation. Our proposed method performs inverse diffusion using deterministic degradation without a model, and the generator of the GAN is trained to perform the diffusion process with the latent random variable. The proposed method uses deterministic degradation, so the latent dimension is low enough to be invertible.
Category: Artificial Intelligence

A Novel Quantum Belief Entropy for Uncertainty Measure in Complex Evidence Theory

Authors: Keming Wu, Fuyuan Xiao
Comments: 2 Pages.

In this paper, a new quantum representation of CBBA is proposed. In addition, a novel quantum belief entropy is proposed to measure the uncertainty of CBBA in complex evidence theory.
Category: Artificial Intelligence

¿cómo Crear Un Pensamiento Y Una Inteligencia Artificial? Y La Matemática De Letras
how to Create an Artificial Thought and Intelligence? And the Math of Letters

Authors: Salvador Sánchez Melgar
Comments: 8 Pages. In Spanish

La construcción de un pensamiento y de una inteligencia artificial es posible con el lenguaje de las letras numeradas. Lenguaje que surgió a través de la creación del libro "Nueva matemáticas de letras, triunfa con la matemática" actualizado con el título "Nueva matemáticas de letras 2ª edición". Libros en los que se exponen el lenguaje de las letras y una matemática de letras donde están las sumas, restas, multiplicaciones y divisiones de letras, con ejemplos y sus correspondientes tablas matemáticas, se podrían hacer con la matemática de las letras cualquier tipo matemático. puesto que es una matemática como la matemática que conocemos.Con el lenguaje de las letras numeradas, que representan letras, palabras y oraciones numeradas, un robot con inteligencia artificial podría adquirir un sin fin de todo tipo de información obtenida por cualquier sentido artificial. Informaciones numéricas que se tendrían que transformar en números binarios.

The construction of a thought and an artificial intelligence is possible with the language of numbered letters. Language that arose through the creation of the book "New mathematics of letters, triumph with mathematics" updated with the title "New mathematics of letters 2nd edition". Books in which the language of letters and a mathematics of letters are exposed where there are additions, subtractions, multiplications and divisions of letters, with examples and their corresponding mathematical tables, any type of mathematics could be done with the mathematics of letters. since it is a mathematics like the mathematics we know.With the language of numbered letters, which stand for letters, words, and numbered sentences, an artificially intelligent robot could acquire endless all kinds of information obtained by any artificial sense. Numeric information that would have to be transformed into binary numbers.
Category: Artificial Intelligence

El Lenguaje De La Inteligencia Artificial Y La Matemática De Letras
the Language of Artificial Intelligence and the Mathematics of Letters

Authors: Salvador Sánchez Melgar
Comments: 27 Pages. In Spanish

Presentación de una matemática de letras y de un lenguaje de letras que le permitirá a una inteligencia artificial aprender sin fin y poder pensar como pensamos nosotros. Con las letras numeradas las informaciones que una inteligencia artificial obtenga con sus sentidos artificiales no perderán sus significados, puesto que mediante estas letras las informaciones se podrán transformar en palabras y numeradas. Cada información que una inteligencia artificial obtenga, la podrá transformar en números binarios, luego en números ordinarios de las letras numeradas, pudiendo así formar palabras numeradas sobre informaciones individuales y globales. Como cada sentido artificial detecta informaciones diferentes, cada sentido crea su propio lenguaje, eso no impide que todas las informaciones se puedan transformar en números. Las palabras numeradas que se puedan formar con las transformaciones de las informaciones también deberán enlazarse con otras palabras numeradas semejantes indexadas en un diccionario de palabras numeradas, para que así el robot pueda saber el significado de cada información. También a este robot se le debería añadir un programa que le permita entender las uniones de palabras. Con las letras numeradas la información que reciba un robot la podrá transformar en palabras numeradas y así poder memorizarlas permanentemente pudiendo así obtener ilimitada sabiduría. Mediante números binarios obtenidos de las informaciones de todo enlazados a informaciones binarias memorizadas de manera positiva y negativa es como pensamos nosotros. También expondré, con tablas y ejemplos, las sumas, restas, multiplicaciones y divisiones de las letras y un sistema numeral de letras del 0 al 27.

Presentation of a mathematics of letters and a language of letters that will allow an artificial intelligence to learn endlessly and be able to think as we think. With the numbered letters, the information that an artificial intelligence obtains with its artificial senses will not lose its meaning, since through these letters the information can be transformed into words and numbered. Each piece of information that an artificial intelligence obtains can be transformed into binary numbers, then into ordinary numbers of numbered letters, thus being able to form numbered words on individual and global information. Since each artificial sense detects different information, each sense creates its own language, this does not prevent all information from being transformed into numbers. The numbered words that can be formed with the transformations of the information must also be linked to other similar numbered words indexed in a dictionary of numbered words, so that the robot can know the meaning of each information. A program should also be added to this robot that allows it to understand word unions. With the numbered letters, the information that a robot receives can be transformed into numbered words and thus be able to memorize them permanently, thus being able to obtain unlimited wisdom. Through binary numbers obtained from the information of everything linked to binary information memorized in a positive and negative way is how we think. I will also expose, with tables and examples, the addition, subtraction, multiplication and division of the letters and a number system of letters from 0 to 27.
Category: Artificial Intelligence

Neuro-symbolic Meta Reinforcement Learning for Trading

Authors: S. I. Harini, Gautam Shroff, Ashwin Srinivasan, Prayushi Faldu, Lovekesh Vig
Comments: 4 Pages. Accepted at Muffin@AAAI'23

We model short-duration (e.g. day) trading in financial mar- kets as a sequential decision-making problem under uncer- tainty, with the added complication of continual concept- drift. We therefore employ meta reinforcement learning via the RL2 algorithm. It is also known that human traders often rely on frequently occurring symbolic patterns in price series. We employ logical program induction to discover symbolic patterns that occur frequently as well as recently, and ex- plore whether using such features improves the performance of our meta reinforcement learning algorithm. We report ex- periments on real data indicating that meta-RL is better than vanilla RL and also benefits from learned symbolic features.
Category: Artificial Intelligence

A General Theory of Artificial Intelligence Part 2

Authors: Matthew Groom
Comments: 6 Pages.

Where to start in growing a real Artificial Intelligence. Let us begin building the first AI, in this paper I will theoretically build an AI from scratch, so I will go through what to do, where to do it.
Category: Artificial Intelligence