• ## Gradient descent method for machine learning

Time：2021-10-25

Suppose we have time and computational resources to calculate the loss of all possible values of W1. For the regression problem we have been studying, the graph of the loss and W1 is always convex. In other words, the graph is always a bowl graph, as shown in the following figure: The loss and weight […]

• ## Statistical learning 3: linear support vector machine (SVM)

Time：2021-10-20

learning strategy Soft interval maximization The “linearly separable support vector machine” defined in the previous chapter requires that the training data be linearly separable. However, in practice, the training data often include outliers, so it is often linear and inseparable. This requires us to make some modifications to the algorithm in the previous chapter, that […]

• ## Simplified regularization of machine learning: L2 regularization

Time：2021-10-19

Please see the following generalization curve, which shows the loss of training set and verification set relative to the number of training iterations Figure 1. Loss of training set and verification set Figure 1 shows that the training loss of a model gradually decreases, but the verification loss eventually increases. In other words, the generalization […]

• ## Logistic regression in machine learning: model training

Time：2021-10-14

Loss function of logistic regression The loss function of linear regression is the square loss. The loss function of logistic regression is a logarithmic loss function, which is defined as follows: \displaystyle LogLoss = \sum_{(x,y)\in D} – ylog(y’) – (1-y)log(1-y’) Of which:1. (x, y) \ in D is a dataset containing many labeled samples (x, […]

• ## Plug and play! Rising point artifact in video super score: iseebeter

Time：2021-9-30

Compile | CVReport | I love computer vision (wechat ID: aicvml) CNN makes the super score results more real, and Gan makes the super score results Fuller, so CNN + Gan = good! Adding a discriminator component can increase the result by 0.32db, plug and play, rising point artifact! Whether it is also feasible in […]

• ## Comparison between self supervised comparison loss and supervised comparison loss

Time：2021-9-8

By Samrat SahaCompile VKSource: towards Data Science Supervised contractual learning this paper discusses a lot between supervised learning, cross entropy loss and supervised contrast loss, so as to better realize the task of image representation and classification. Let’s take a closer look at the content of this paper. The paper points out that there can […]

• ## Xmake v2.3.2 is released, bringing the same construction speed as ninja

Time：2021-8-13

This version focuses on refactoring and optimizing the internal parallel construction mechanism, realizing the parallel compilation of source files between multiple targets and the support of parallel links. At the same time, it optimizes some internal losses of xmake and fixes some bugs affecting the compilation speed.Through test and comparison, the current overall construction speed […]

• ## Extension end data tecdat: Python value at risk calculation Portfolio VaR (value at risk), expected loss es

Time：2021-8-10

Original link:http://tecdat.cn/?p=22788 Python calculates the risk measure of a multi asset portfolio. Key concepts As the price changes, the market value held by the investment manager will also change. The latter is the so-called market risk. One of the most popular methods to measure it is defined as value at risk. Risk itself is seen […]

• ## Machine learning (2): understand linear regression and gradient descent and make simple prediction

Time：2021-8-3

Prediction begins with guesswork PressLast articleMachine learning is the process of applying mathematical methods to find laws in data. Since mathematics is the interpretation of the real world, let’s return to the real world and make some comparative imagination. Imagine that there is a white board made of plastic foam in front of us. There […]

• ## Target detection in the loss of focus of the introductory guide!

Time：2021-7-21

By guest blogCompile FlinSource: analyticsvidhya introduce Object detection is one of the most widely studied topics in the computer vision community. It has entered various industries, involving use cases from image security, surveillance, automatic vehicle systems to machine inspection. At present, object detection based on deep learning can be roughly divided into two categories Two […]

• ## Densebox: early anchor free research with advanced ideas | CVPR 2015

Time：2021-7-13

The design of densebox detection algorithm is very advanced. Nowadays, many anchor free methods have their own shadow. If it didn’t appear a little later than fast r-cnn at that time, the field of target detection might have been developing in the direction of anchor free for a long time  Source: Xiaofei’s algorithm Engineering Notes […]

• ## Tecdat: Python Monte Carlo simulation of portfolio value at risk (VaR)

Time：2021-7-8

Link to the original text:http://tecdat.cn/?p=22862  How to use Python to automatically calculate the value at risk (VaR) through Monte Carlo simulation to manage the financial risk of portfolio or stock. VaR in financial and portfolio risk management? VaR is the abbreviation of “value at risk”, which is used by many companies and banks to determine […]