Category:Artificial Intelligence

  • The LTP model of Harbin Institute of Technology


    1、 Installation 1. I spent a whole morning and a half in the afternoon installing. Finally, I realized that LTP is different from pyltp, and the mature version is (that’s what it is) I understand that some projects of GitHub can also view the historical release. The serious 3.4.0 is found in the releaseComplexity: […]

  • Simple classification of deep learning


    Simple classification of deep learning Simple binary classification Manufacturing data from sklearn.model_selection import train_test_split from sklearn import datasets import matplotlib.pyplot as plt from tensorflow import keras X,y = datasets.make_blobs(n_samples=1000,random_state=8,centers=2) plt.scatter(X[:,0],X[:,1],c=y) Building models and training model = keras.models.Sequential([ keras.layers.Dense(32,input_shape=X.shape[1:]), keras.layers.Dense(1,activation=keras.activations.sigmoid)] ) model.summary() model.compile(loss = keras.losses.binary_crossentropy,optimizer = keras.optimizers.RMSprop(learning_rate=0.1),metrics = [keras.metrics.Accuracy()]),y,validation_split=0.25,epochs = 20) View test data […]

  • The tail is processed as a time series to identify whales


    By Lamothe ThibaudCompile FlinSource: to ward data science Using curvature integral and dynamic time warping, let’s study sperm whale recognition in depth! preface Recently, we tried Capgemini’s global data science challenge. I worked with ACORES whale research center to identify sperm whales and use AI to help save their lives. To accomplish this task, we […]

  • Performance Optimization: thread resource recycling


    This article is from:Perfma technology community Perfma official website 1、 Questions There are many timeouts in the sorting request of the model service platform, and the null pointer exception is accompanied from time to time. 2、 Changes before and after the problem The recall engine expands the recall amount, resulting in an increase in the […]

  • Watermark in Flink


    watermark Premise of use: The time semantics of processing data is event time, that is, the time when each data is generated. Usage scenarios (problem solving) Dealing with out of order data: in Flink, data is processed in real time. However, due to the problem of network transmission, the data generated first will arrive later, […]

  • Kafka producer messaging design


    Previous articles analyzed the sending process of Kafka and the use of NiO, but there are still many holes left. Here is a summary of the remaining problems. Why should the received data be cached? When the selector in Kafka reads the data from the remote end, it will cache the received data first private […]

  • Deep reinforcement learning


    lecture 1 policy gradient Actor makes a decision action according to the environment state, and gets a reward after the decision Once a game becomes an epic, calculate the total reward of an epic trajectory tao = { s1,a1,s2,a2,….,sT,aT} P theta (tao) = p(s1)ptheta(a1|s1)p(s2|s1,a1)theta(a2|s2)… R (TAO) = total reward in one expectation gradient ascent Gradient […]

  • Analysis of kmeans in spark


    Analyze the kmeans code, which is a little more complicated import numpy as np from pyspark import SparkContext #The purpose of this function is to convert the read data into float data def parseVector(line): return np.array([float(x) for x in line.split(‘ ‘)]) #The purpose of this function is to find out which point set the point […]

  • Cornernet: Classic keypoint based method for target detection by locating corners | eccv2018


    In this paper, cornernet is proposed to detect targets by detecting corner pairs, which has the same performance as the current SOTA detection model. Cornernet uses the method of human pose estimation for reference, and creates a new framework in the field of target detection. Many papers based on corernet develop new corner target detection  […]

  • R language Stan for Bayesian reasoning analysis


    Link to the original text: Stan of R Stan can be run from many statistical packages. So far, I’ve been running Stan from R. Simple linear regression The first step is to document the Stan model. This contains a file linreg.stan : data { int N; vector[N] x; vector[N] y; } model { y ~ […]

  • Installing Chinese font on Linux


    cause Recently, I want to use Python Matplotlib to draw pictures on raspberry pie, but I find that it can’t display Chinese. A specified font found that simhei font is not installed. Solution Copy from windows simhei.ttf Font file to raspberry pie (I saved a copy on GitHub:…)To create a new folder. Then refresh the […]

  • Building dockers on raspberry pie


    cause I’ve always heard people say that it’s better to use docker in raspberry pie. Recently, I made a deduction and found that it’s really good, at least the isolation effect is very good. You know, because raspberry pie is based on ARM architecture, more PIP packages can only be installed in the system in […]