QA Official

Text Mining Research

https://qaofficial.com/post/2019/04/21/70563-text-mining-research.html 2019-04-21
1. Text Mining Definition Text mining refers to the discovery of hidden patterns P from the set C of large amounts of text.If C is regarded as input and P as output, then the process of text mining is a mapping F: C-P from input to output.It is a process of obtaining interesting or useful patterns from text information. 2. Development of Text Mining The traditional information retrieval technology is

Text Processing and Feature Engineering in Natural Language Processing

https://qaofficial.com/post/2019/04/21/70555-text-processing-and-feature-engineering-in-natural-language-processing.html 2019-04-21
Almost Human Report Text Processing In the existing data, text is the most unstructured form, with various noises inside;Without preprocessing, text data cannot be analyzed.The whole process of cleaning and standardizing text is called textpreprocessing. Its function is to make text data noise-free and can be analyzed. mainly includes three steps: noise removal vocabulary normalization object standardization The following figure shows the structure of the text preprocessing process. Remove noise

Text vectorization

https://qaofficial.com/post/2019/04/21/70553-text-vectorization.html 2019-04-21
Express the text into a form that can be understood by the computer. The so-called text expression refers to the vectorization of the text.Text vectorization can be divided into vector expression of words, vector expression of short text and vector expression of long text, because different situations require different methods and processing methods. Ignoring these details, I made a survey on vectorization of texts under normal circumstances. Common ideas are as follows:

The Research Progress and Prospect ofDepth Learning in Image Recognition

https://qaofficial.com/post/2019/04/21/70646-the-research-progress-and-prospect-ofdepth-learning-in-image-recognition.html 2019-04-21
The Research Progress and Prospect ofDepth Learning in Image Recognition deep learning is one of the most important breakthroughs in the field of artificial intelligence in the past decade.It has achieved great success in many fields such as speech recognition, natural language processing, computer vision, image and video analysis, multimedia, etc.This article will focus on the latest research progress of depth learning in object recognition, object detection and video analysis, and discuss its development trend.

Transfer of learning and Fine-tuning Depth Convolutional Neural Network

https://qaofficial.com/post/2019/04/21/70644-transfer-of-learning-and-fine-tuning-depth-convolutional-neural-network.html 2019-04-21
Original Link: https://blog.csdn.net/bigbzheng/article/details/52372946(This article only selects the knowledge points that I am not familiar with for labeling) Keywords: Depth Convolutional Neural Network (DCNN), Transfer of learning, fine-tuning Current research has proved that DCNN is very effective in automatically analyzing a large number of images and identifying image features, and it can minimize the error rate of image classification.DCNN seldom trains from scratch because it is not so easy to obtain a set of data in a specific field with a large enough sample.

[Model Selection and Evaluation 03] Model Evaluation: Quantifying the Quality of Forecast

https://qaofficial.com/post/2019/04/21/70502-model-selection-and-evaluation-03-model-evaluation-quantifying-the-quality-of-forecast.html 2019-04-21
1. ReferencesSklearn document There are 3 different API to evaluate the quality of model predictions: [1] estimator score method: Estimators have a score method, which provides default evaluation criterion for the problems they solve.There are no related discussions on this page, but there are related discussions in each estimator document. [2] scoring Parameter: Model-evaluation tools relies on internal scoring strategy using cross-validation (such as Model _ Selection. Cross _ Val

aself-pacific multiple-instance learning framework for co-saliency detection

https://qaofficial.com/post/2019/04/21/70531-aself-pacific-multiple-instance-learning-framework-for-co-saliency-detection.html 2019-04-21
Summary:The traditional co-saliency is to extract artificial feature matrix, which lacks the ability to generalize to various scenes.Moreover, there is a lack of biological mechanism theory. In order to solve this problem, a novel task is proposed.Through multi-instance learning and self-regulated learning.On the one hand, it is suitable for matrix measurement method (evaluation method). It is found that the common part passes MIL in the co-salient region in the self-learning method.

error2203 encountered while installing vm

https://qaofficial.com/post/2019/04/21/70616-error2203-encountered-while-installing-vm.html 2019-04-21
error2203 encountered while installing vm.The following methods can be found and solved: We may get this error due to incorrect value of TEMP/TMP environment variable. You may perform the following steps to get the issue resolved. 1. Goto My Computer->Properties->Advanced->Environment Variables. 2. Select TEMP System Variable and Click Edit. 3. Validate the path. Set it to a valid folder on the local machine.checked my environment variables and found them in the system variables respectively.

machine learning: Colorization using Optimization

https://qaofficial.com/post/2019/04/21/70672-machine-learning-colorization-using-optimization.html 2019-04-21
Today, we introduce an article by Siggraph 2004: Colorization using Optimization, which uses the optimization method to color gray images. Here, we use the very classical Poisson equation and the linear optimization of sparse matrices.In a nutshell, it is to color a gray-scale image first, and then fill other areas without color with an optimized method.These processes are all carried out in YUV color space. Given a Y-channel image, we hope to restore the U and V channels of the image based on certain prior knowledge.

10g Perform Database Import and Export with sys User

https://qaofficial.com/post/2019/04/20/68433-10g-perform-database-import-and-export-with-sys-user.html 2019-04-20
Error Phenomena: [[email protected] data]$ exp "sys/[email protected] as sysdba" file=/data/sys.dmp log=/data/sys.log full=y; LRM-00108: invalid positional parameter value 'as' EXP-00019: failed to process parameters, type 'EXP HELP=Y' for help EXP-00000: Export terminated unsuccessfully Reason: This may be a BUG in Oracle10g, just follow the following two methods: Solution: first method: [ Oracle @ wwldata ] $ exp \ " sys/Oracle assysdba \" file=/data/sys.dmp log=/data/sys.log full=y; Export: Release 10.2.0.1.0 - Production on Wed May 2 19:20:12 2012