• Lightweight Network Overview – backbone network


    The core of lightweight network is the lightweight transformation of the network from both volume and speed on the premise of maintaining the accuracy as much as possible. This paper briefly describes the lightweight network, mainly involving the following networks: Squeezenet series Shufflenet series MnasNet Mobilenet series CondenseNet Espnet series ChannelNets PeleeNet IGC series Fbnet […]

  • Hire MLP of vision MLP: vision MLP via hierarchical arrangement


    Hire-MLP: Vision MLP via Hierarchical Rearrangement Original document:https://www.yuque.com/lart/pa… This article is very easy to read. There are no complex words and sentence patterns. It is very smooth from beginning to end. I like this writing style very much. Learn about the article from the summary This paper presents Hire-MLP, a simple yet competitive vision MLP […]

  • [openmldb meetup #1] meeting minutes


    1. Meeting contents On January 15, 2022, the openmldb community held the first meetup for the whole community. The core development team of openmldb not only shared the overall architecture and v0 4.0, and invited akulaku, an enterprise customer of openmldb, to share the actual combat scenario of real-time feature calculation based on openmldb. The […]

  • Image segmentation with aggregation


    Dense Prediction with Attentive Feature Aggregation Original document:https://www.yuque.com/lart/pa… The paper accidentally turned over from arXiv can be regarded as an extension of the previous work Read the paper from the abstract Aggregating information from features across different layers is an essential operation for dense prediction models. This paper focuses on the integration with cross – […]

  • Cycle MLP a MLP like architecture for dense prediction of visoin MLP


    Cycle MLP a MLP like architecture for dense prediction of visoin MLP Original document:https://www.yuque.com/lart/pa… Read the article from the abstract This paper presents a simple MLP-like architecture, CycleMLP, which is a versatile backbone for visual recognition and dense predictions, unlike modern MLP architectures, e.g., MLP-Mixer, ResMLP, and gMLP, whose architectures are correlated to image size […]

  • Feature Engineering in machine learning (I) — Overview


    preface Last month, I participated in the wechat big data challenge. Because it was the first time to participate in a similar competition, I didn’t have any experience and didn’t enter the semi-finals in the end. However, during this period, I still learned a lot of knowledge, especially the content related to feature processing. For […]

  • Practice of the fourth paradigm openmldb in recalculation optimization of financial risk control database


    Recently, at the datafunsummit: Intelligent Finance Online summit, Chen Dihao, the fourth paradigm platform architect, focused on the application of the fourth paradigm open source machine learning database openmldb in the financial field, as well as the processing of underlying chronological features and the details of window calculation optimization, with the theme of openmldb risk […]

  • Can Cerna pure student letters still be sent for 5 points now? Come and watch


    Title: comprehensive analysis of Cerna network to explore the prognostic characteristics of renal papillary cell carcinoma (kirp) The idea of establishing a disease-related Cerna network is roughly as follows: first screen the lncrna (circrna), miRNA and mRNA that are significantly differentially expressed in the disease, and then predict the interaction between them, so as to […]

  • [R language] chi square test and Fisher exact test, reproducing clinical paper


    When doing clinical data analysis, we often use chi square test or Fisher exact test to see if there are significant differences in different clinical characteristics between the two groups. Today, Xiaobian will take you to reproduce Table 2 of the following paper Table 2 mainly shows whether there are significant differences in various clinical […]

  • Machine learning algorithm series (XIII) – naive Bayes classifier algorithm


    Background knowledge required for reading this article: first, lose programming knowledge 1、 Introduction    the previous sections introduce one kind of classification algorithm – linear discriminant analysis and quadratic discriminant analysis, and then introduce another kind of classification algorithm——Naive Bayesian classification algorithm1(Naive Bayes Classifier Algorithm/NB)。 Naive Bayesian classification algorithm is applied in the field of […]

  • Deep learning notes 01 – Introduction


    1 Introduction catalogue 1 Introduction 1.2 machine learning 1.3 express learning Local representation Distributed representation Express learning Traditional feature extraction and representation learning 1.4 deep learning 1.5 artificial neural network 1.2 machine learning Traditional machine learning can be regarded as shallow learning: it does not involve feature learning, and features are extracted by manual experience […]

  • Random forest in ensemble learning


    abstract Random forest is one of the most advanced representatives of ensemble algorithms. Random forest is the upgrade of bagging. The main difference between it and bagging is the introduction of random feature selection. That is, when selecting segmentation points for each decision tree, the random forest will first randomly select a feature subset, and […]