7 results about "Probabilistic framework" patented technology

This invention discloses a crowdsourcing testing platform system and method based on intelligent matching. The method includes: receiving test task requests containing task description text; using a hierarchical attention network model to perform semantic understanding of the task description text, generating structured skill tags, and dynamically updating the tag confidence of testers using a Bayesian probabilistic framework; calculating the initial matching degree using a cosine similarity algorithm that incorporates a recent task completion quality correction factor, and then filtering and constructing a candidate pool; for each candidate, constructing a dynamic feature matrix, and using an improved K-nearest neighbor algorithm based on Mahalanobis distance for ranking and recommendation, where the K value is dynamically determined based on the number of candidates in the pool and the task dwell time. This invention significantly improves the efficiency and accuracy of task allocation through a capability assessment mechanism, achieving optimized allocation of testing resources.

Probabilistic models may be used in a deterministic workflow for an automation. Artificial intelligence (AI) introduces a probabilistic framework where the outcome is not deterministic, and therefore, the steps are not deterministic. Deterministic workflows may be mixed with probabilistic workflows, or probabilistic activities may be inserted into deterministic workflows, in order to create more dynamic workflows. A supervisor system may be used to monitor an AI model and raise an alarm, disable an automation, bypass the automation, or roll back to a previous version of the AI model when an error is detected by a data drift detector, a concept drift detector, or both.

A precursor ion transmission window is moved in overlapping steps across a precursor ion mass range. The precursor ions transmitted at each overlapping step by the mass filter are fragmented or transmitted. Intensities or counts are detected for each of the one or more resulting product ions or precursor ions for each overlapping window that form mass spectrum data for each overlapping window. Each unique product ion detected is encoded in real-time during data acquisition. This encoding includes sums of counts or intensities of each unique ion detected the overlapping windows and positions of the windows associated with each sum. The encoding for each unique ion is stored in a memory device rather than the mass spectral data. A deblurring algorithm or numerical method is used to determine a precursor ion of each unique ion from the encoded data.

A precursor ion transmission window is moved in overlapping steps across a precursor ion mass range. The precursor ions transmitted at each overlapping step by the mass filter are fragmented or transmitted. Intensities or counts are detected for each of the one or more resulting product ions or precursor ions for each overlapping window that form mass spectrum data for each overlapping window. Each unique product ion detected is encoded in real-time during data acquisition. This encoding includes sums of counts or intensities of each unique ion detected the overlapping windows and positions of the windows associated with each sum. The encoding for each unique ion is stored in a memory device rather than the mass spectral data. A deblurring algorithm or numerical method is used to determine a precursor ion of each unique ion from the encoded data.

The invention discloses a process industry system simulation method based on a state space model, solves the problem of inaccurate parameter estimation of a process industry system model under unknown time lag and non-Gaussian interference noise, and belongs to the technical field of process industry system modeling. The method comprises the following steps: establishing a linear time-delay state space model for simulating a process industrial system; under a probability framework, polynomial distribution is introduced to model random time delay between a state variable and an output variable of the time delay state space model, Dirichlet distribution is utilized to describe time delay prior information, and probability distribution of unknown parameters in the time delay state space model is determined; and based on the variational Bayesian reasoning framework and the probability distribution of the unknown parameters, identifying the unknown parameters in the linear time-delay state space model according to the existing process industrial system control quantity and the sensor acquisition quantity, and obtaining the posterior distribution of the unknown parameters. According to the method, accurate modeling of the process industrial system under the high-dynamic and multi-interference conditions can be realized.

The invention provides a thinking chain selection method based on internal confidence of a large language model, and belongs to the field of artificial intelligence. Relates to a big language model reasoning optimization technology, in particular to a training-free thinking chain selection method based on big language model inherent confidence, and aims to improve reasoning efficiency and evaluate reasoning process rationality and answer accuracy at the same time. Based on a probability framework and taking internal confidence as a scoring signal, setting selection of an optimal answer question as maximum likelihood estimation for solving a joint probability distribution space of a thinking chain and a final answer; setting X to represent a possible prompt set, R to represent a thinking chain set and Y to represent a possible final answer set, and for a given input prompt x belongs to X, an optimal thinking chain r belongs to R and an answer y belongs to Y corresponding to the optimal thinking chain; identifying a pair (r, y) with the highest joint conditional probability; wherein the probability is a joint conditional probability, and refers to the probability that the model generates a thinking chain r and a final answer y under a given prompt x.

Probabilistic models can be used in deterministic workflows for robotic process automation (RPA). Machine learning (ML) introduces a probabilistic framework where the outcome is uncertain, and thus the steps are uncertain. Deterministic workflows can be mixed with probabilistic workflows, or probabilistic activities can be inserted into deterministic workflows in order to create more dynamic workflows. A governance system can be used to monitor ML models and issue alerts, disable RPA robots, bypass RPA robots, or roll back to a previous version of the ML model when an error is detected by a data drift detector, a concept drift detector, or both.