241:
320:
118:
22:
63:
291:
There are many hyperplanes that might classify the data. One reasonable choice as the best hyperplane is the one that represents the largest separation, or margin, between the two classes. So we choose the hyperplane so that the distance from it to the nearest data point on each side is maximized. If
81:
139:
35:
361:
288:) as to why maximizing the margin (under some suitable constraints) may be beneficial for machine learning and statistical inferences algorithms.
385:
190:
227:
209:
162:
99:
49:
380:
169:
143:
41:
354:
276:. Note that there are many distances and decision boundaries that may be appropriate for certain datasets and goals. A
176:
128:
158:
147:
132:
294:
347:
281:
327:
183:
240:
301:
277:
273:
331:
265:
374:
280:
is a classifier that explicitly utilizes the margin of each example while learning a
285:
117:
319:
307:
272:
of a single data point is defined to be the distance from the data point to a
239:
111:
56:
15:
335:
77:
298:and the linear classifier it defines is known as a
284:There are theoretical justifications (based on the
72:
may be too technical for most readers to understand
355:
292:such a hyperplane exists, it is known as the
8:
146:. Unsourced material may be challenged and
50:Learn how and when to remove these messages
362:
348:
228:Learn how and when to remove this message
210:Learn how and when to remove this message
100:Learn how and when to remove this message
84:, without removing the technical details.
260:separates them with the maximum margin.
82:make it understandable to non-experts
7:
316:
314:
144:adding citations to reliable sources
254:does, but only with a small margin.
334:. You can help Knowledge (XXG) by
14:
31:This article has multiple issues.
318:
116:
61:
20:
39:or discuss these issues on the
248:does not separate the classes.
159:"Margin" machine learning
1:
402:
386:Computer programming stubs
313:
295:maximum-margin hyperplane
381:Support vector machines
305:; or equivalently, the
330:-related article is a
261:
310:of optimal stability.
243:
328:computer-programming
140:improve this article
262:
343:
342:
302:margin classifier
278:margin classifier
274:decision boundary
238:
237:
230:
220:
219:
212:
194:
110:
109:
102:
54:
393:
364:
357:
350:
322:
315:
266:machine learning
233:
226:
215:
208:
204:
201:
195:
193:
152:
120:
112:
105:
98:
94:
91:
85:
65:
64:
57:
46:
24:
23:
16:
401:
400:
396:
395:
394:
392:
391:
390:
371:
370:
369:
368:
259:
255:
253:
249:
247:
234:
223:
222:
221:
216:
205:
199:
196:
153:
151:
137:
121:
106:
95:
89:
86:
78:help improve it
75:
66:
62:
25:
21:
12:
11:
5:
399:
397:
389:
388:
383:
373:
372:
367:
366:
359:
352:
344:
341:
340:
323:
257:
251:
245:
236:
235:
218:
217:
124:
122:
115:
108:
107:
90:September 2010
69:
67:
60:
55:
29:
28:
26:
19:
13:
10:
9:
6:
4:
3:
2:
398:
387:
384:
382:
379:
378:
376:
365:
360:
358:
353:
351:
346:
345:
339:
337:
333:
329:
324:
321:
317:
312:
311:
309:
304:
303:
297:
296:
289:
287:
283:
279:
275:
271:
267:
242:
232:
229:
214:
211:
203:
192:
189:
185:
182:
178:
175:
171:
168:
164:
161: –
160:
156:
155:Find sources:
149:
145:
141:
135:
134:
130:
125:This article
123:
119:
114:
113:
104:
101:
93:
83:
79:
73:
70:This article
68:
59:
58:
53:
51:
44:
43:
38:
37:
32:
27:
18:
17:
336:expanding it
325:
306:
299:
293:
290:
286:VC dimension
269:
263:
224:
206:
197:
187:
180:
173:
166:
154:
138:Please help
126:
96:
87:
71:
47:
40:
34:
33:Please help
30:
282:classifier.
200:August 2014
375:Categories
308:perceptron
170:newspapers
36:improve it
127:does not
42:talk page
300:maximum
184:scholar
148:removed
133:sources
76:Please
270:margin
186:
179:
172:
165:
157:
326:This
191:JSTOR
177:books
332:stub
268:the
163:news
131:any
129:cite
264:In
142:by
80:to
377::
45:.
363:e
356:t
349:v
338:.
258:3
256:H
252:2
250:H
246:1
244:H
231:)
225:(
213:)
207:(
202:)
198:(
188:·
181:·
174:·
167:·
150:.
136:.
103:)
97:(
92:)
88:(
74:.
52:)
48:(
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.