Vehicle financial loan waiting time estimation method, device, and calculation medium
A waiting time and financial technology, applied in finance, computing, neural learning methods, etc., can solve problems such as poor prediction accuracy, impact of waiting time, large difference in processing time, etc., to avoid data errors, improve prediction accuracy, The effect of reducing the burden
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0041] This embodiment proposes a method for estimating the waiting time of an auto finance loan based on a neural network, including a data collection and processing stage, a model training stage, and a prediction stage.
[0042] In the data collection and processing stage, it is first necessary to collect multiple auto finance loan application forms, extract and encode the characteristic information of each auto finance loan application form. In this embodiment, data related to queuing time for about half a year was collected by burying points on the big data platform, including policy data, personnel data, process time node data, etc. Specifically, the data extracted from the big data platform mainly consists of four tables: Application form processing flow table, application form level table, application form level rule table and approver level rule table, enter simple data processing such as table connection, and the final production fields include: application form number...
Embodiment 2
[0048] This embodiment is improved on the basis of Embodiment 1, and a specific method for extracting time features is proposed, but it is obvious that the time feature extraction method of the present invention is not limited to the form of this embodiment.
[0049] The time feature can include the entry queue time, processing start time, and processing completion time of the auto finance loan application form. Taking the extraction of the entry queue time as an example, the main elements of the time feature are the time when the application form entered the queue, the day of the week, and the time period. In this embodiment, considering business commuting and rest time (such as going to work at 8.30, lunch break at 11.30-1.30), 30 minutes are divided into a time period, and the form of the first few minutes in the time period is reflected, so that the expression form of the extracted time feature It is (a, b, c), where a, b and c are all positive integers, and a∈[1, 7] repres...
Embodiment 3
[0052] This embodiment improves on the basis of Embodiments 1 and 2, and proposes a specific method for encoding time features. In this embodiment, the time features are average-coded, and the formula is as follows:
[0053]
[0054] Where x is the time feature to be encoded. For example, when encoding the time feature extracted in Embodiment 2, x involves the three fields a, b, and c, and xi represents the time feature of a certain actual value, such as implementing (3, 21, 13) in Example 2. no i is x=x i When the number of samples, N is the total number of samples, y is the target value, means x=x i The mean value of y corresponding to , is the mean of y over the entire training set.
[0055] λ(n i ) represents the balance coefficient between the special case and the whole, λ(n i )∈[0,1] is responsible for calculating the reliability of two probability values, λ(n i )=0.5 means that the reliability of the two probabilities is equal, with n i As the value increa...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com