# HG changeset patch # User Takahiro SHIMIZU # Date 1555573243 -32400 # Node ID 1b00d07674d740bc883accaf5b606c2668c45edf add note diff -r 000000000000 -r 1b00d07674d7 TR_W/.Advanced software.png Binary file TR_W/.Advanced software.png has changed diff -r 000000000000 -r 1b00d07674d7 TR_W/4_17.md --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/TR_W/4_17.md Thu Apr 18 16:40:43 2019 +0900 @@ -0,0 +1,43 @@ +- Reports Types + - techinal report : engineering magazine + - research report : journals(published regularly) + +1. techinal reports + - a + - writtern by engineers(techinal reports) + - b + - writtern by progessional journal(techical articles) +a) techinal reports can be more difficult to read(words techinal). It is like a text book ( the topics are more new one) +b) techinal articles are less difficult to read (stories in a magazine...) + +2. research reports +- more difficult , has nevelit, has specific style +- are more difficults than techinal report(a) or techical articles(b) +- words are more sientific, follows one formore style (format) +- giveing a new idea for an old method(please give merit) + +++++ +- introduction: idea of others about topic +- methods: the method by authors are explained +- results: explain about using the methods +++++ + + +# formed style + - a) introduction + - b) methods + - c) results + - d) conslusion + - e) reference + (Documentation) + +# words +- reputations + - 有名, 評判 +- progress +- civil + - 市民 + - 国内 + - 民間の +- perhaps +- undoubtedly diff -r 000000000000 -r 1b00d07674d7 TR_W/Advanced software.mm --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/TR_W/Advanced software.mm Thu Apr 18 16:40:43 2019 +0900 @@ -0,0 +1,80 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff -r 000000000000 -r 1b00d07674d7 data_mining_theory/4_16.md --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/data_mining_theory/4_16.md Thu Apr 18 16:40:43 2019 +0900 @@ -0,0 +1,51 @@ +- We can turn data into knowledge + + +- Superviser Algorithm use Labels, detail corpus + +- regression + - 回帰 +- classification + +- Preprocession + - 準備 + - transforming the data +- Learning +- Evaluation + - use test data +- Prediction + - use unknown data + +# WhyPython? + +- Numpy +- Pandas +- Scikit learn + +# QA + +## What's feature many learning costs? + +- Could give us example test regression task? + - surrary? + +# Chap2 + +## Perceptron Algorith +- Classfication algorithm that cased on MCP neuron model + - vectors or input and weight + - use binary data +- w .. wight +- x .. vector data + +# calculate + +z = `\sum{w_i*x_i}` + + +# 英語 +- regression +- classification +- increasing [増加する] + - His artworks are likely to increase in value. + - The number of people over the age of 65 increased to nrearly 23 millions in 2011.