Topic 6: product inspection system
Basic requirements: 1.
[1] is required for a certain types of products in the factory according to the number of attributes to determine whether the product is qualified (one said the product qualified, 0 indicates the product is unqualified), the classification model is as follows:


The x1, x2 for the product attributes, a0, a1, a2 model parameters for the classification, need you now according to the historical data of the product create a binary classification model, after the model according to the two attributes of the product to determine whether a product to be detected qualified (note: subject data given by the end of the article, and the data is produced by the above linear model, establish classification model, you can identify the parameters of the linear model, or you can use other ways of modeling, the ultimate goal is according to the product of two property values to judge the product is unqualified, the establishment of the classification model must be made within the system)

[2] products include the following information: product number, product of at least four attribute value (attribute value set itself), such as
2. The basic management functions:
[1] product information input: input from the keyboard product two attribute values and the number of
[2] products the preservation of historical information and read: should be saved to the hard disk (in the form of the database file or plain text files), when need can be read from the hard disk data will come out,
[3] product information, delete and modify, delete this number of products, or modify its attribute value and number
[4] classification model: according to the historical data provided by classification model, this part of content need to consult the parameter identification of relevant information or knowledge such as machine learning related concepts, and programming),
[5] product inspection: according to the established model, the two attributes of the product value as input, whether it is qualified
[6] parameter identification results evaluation: according to the model of binary classification model is set up by the judgment results, get a misjudgment rate for products;
[7] information preservation: after testing the product, to save the information of the product
[8] print function: print test unqualified products information
[9] system encourages people to design more functionality,

Product number product attribute x1 x2 qualified product attributes with
Product the historical data of 0.999355 0.841053 100001 1
100002 0.601533 0.169247 1
100003-0.354169-0.970955 0
100004-0.949711-0.685297 0
100005 0.465647 0.895195 1
100006-0.054427-0.455036 0
100007-0.214823-0.770084 0
100008 0.720028 0.029238 1
100009-0.885077-0.737189 0
100010-0.230910-0.966592 0
100011-0.716319-0.637817 0
100012-0.908637-0.567747 0
100013 0.395989 0.612585 1
100014 0.827023 0.825689 1
100015-0.232909-0.367380 0
100016-0.513778-0.776787 0
100017 0.814195 0.425739 1
100018 0.404036 0.030216 1
100019 0.791276 0.666436 1
100020 0.895947 0.818507 1
100021 0.873384 0.949016 1
100022 0.993573 0.328544 1
100023-0.206115-0.026155 0
0.319873 0.549274 0, 100024
100025 0.645901 0.109418 1
100026 0.244733 0.362297 1
100027 0.401013 0.669886 1
100028-0.020970-0.791746 0
0.017634 0.124804 0, 100029
100030 0.559427 0.200689 1
100031-0.919028-0.119760 0
100032 0.802015 0.380747 1
100033 0.464091 0.106093 1
100034 0.231566 0.928780 1
100035 0.510376 0.075304 1
100036-0.431066-0.008016 0
100037-0.587502-0.956920 0
0.304372 0.615430 0, 100038
100039 0.856240 0.796269 1
100040 0.609147 0.221416 1
100041-0.299128-0.795908 0
100042-0.459623-0.300603 0
100043 0.489235 0.642908 1
100044 0.445984 0.501818 1
100045 0.713441 0.281969 1
100046 0.837639 0.951508 1
100047-0.461983-0.841328 0
0.376304 0.892386 0, 100048
100049 0.235994 0.806169 1
100050-0.856893-0.620001 0
100051-0.623204-0.985230 0
100052-0.328861-0.347001 0
100053-0.377908-0.544913 0
100054-0.159592-0.887421 0
100055-0.824522-0.621694 0
100056 0.119737 0.848888 1
100057 0.335250 0.437004 1
100058-0.769397-0.758513 0
100059-0.368679-0.730124 0
100060 0.961593 0.285150 1
Product data to be detected 100061 0.771739 0.727447 1
100062-0.954483-0.611561 0
100063 0.726571 0.362250 1
100064-0.510013-0.137226 0
100065-0.132797-0.894231 0
100066-0.729248-0.902345 0
100067-0.568800-0.935839 0
100068 0.549458 0.447666 1
100069-0.825627-0.864494 0
100070-0.199244-0.653899 0
0.180222 0.207353 0, 100071
100072-0.606953-0.162353 0
100073 0.393625 0.312932 1
100074 0.185939 0.942419 1
100075-0.071620-0.206812 0
100076-0.517690-0.335282 0
100077-0.174459-0.726650 0
100078 0.540989 0.159826 1
100079-0.929322-0.853178 0
100080 0.215101 0.742794 1
Note: according to the product of the historical data, we get two classification model, after the red information is assuming unknown, two attributes of data for test sample to judge whether the product is qualified,