2007年11月16日 星期五

ABSTRACT

總算擠出來了... 該好好感謝一下老天爺



但拼死寫這個到底做啥用? 又從中獲得啥東西啊?



研究動機: 畢業...(-.-?)



除此之外想不到其他的啊!  越寫越違背良心..........





本來是交給大眼睛博士, 雖然不到幾秒就生出來



可是, 看沒三句... 憂喜參半



憂的是, 它廠牌給我翻成factory cards......



照這種情況, 繼續看下去恐怕會昏倒;



喜的是, 原來我並不是最弱的...XD





結果到頭來, 眼睛博士跟廢物沒兩樣!



還是得自己花2HR從頭到尾全部重翻



hu... 總算改完了



應該...有...好一點(吧?) -.-;;



頭已經快爆炸...Orz



到了叫我..........





====

   As internet network being developed, optics reads, burns the progress

which records and storage technology, DVD has already become the mainstream

of storing the media at present. There is more and more DVD of different

brands on the market, and quality is widely different. Besides brand, the

looks, relatively in traditional CD, DVD has more diversified nature

functionally in the specification, capacity, the market structure of high

complexity, make consumers while choosing DVD products, may increase the

difficulty of decision.


   In order to consider in terms of consumer this research, through the

experimental design, burning and testing on the spot. And find out the key

factor which may influence DVD's quality of record in various kinds of

factors. Discuss and analyses in testing the income data, and find out the

objective high performance, low cost combination. The result of the testing

will help general consumers choose suitable strategy to purchase DVD based

on the demand of the different levels.


   Secondly, using fundamental categorized rule to separate and sort by

DVD's relative horizontal of quality. Observe different district of clusters

, ensuring the effectiveness of clustering at the same time. And then combine

"prices" to compare and assess synthetically, try to calculate the Capabi-

lity/Price (C/P) ratio of every DVD product in this research. Moreover, im-

plement Support Vector Machines Regression (SVMR) to analyze the trend of

quality decline of record. Offer users to make follow-up track in connection

with the decline of DVD quality phenomenon. Using this tracking mechanism

can help users to back up early before DVD quality damage, and reduce risk

of data loss.





【原始人版】

   As internet network being developed, optics reads, burns the progress

which records and stores technology, DVD has already become the mainstream

of storing the media at present . There is more and more DVD of different

factory cards on the market, quality is unavoidable and different . Besides

factory card, the looks, relatively in traditional CD, DVD and has more

diversified nature functionally in the specification , capacity, the market

structure of high complexity, make consumers while choosing DVD products,

difficulty degree that also increases much decision and has.


   In order to consider in terms of consumer this research, through the

experimental design, burn and record testing actually, it may influence

DVD to burn the key factor which record quality to find out various kinds

of , discuss that analyses in testing the income data, and find out the

objective high performance, low cost association , the result of the income

contributes to general consumers, to the demand of the different levels ,

choose suitable DVD to purchase the tactics.


   Secondly, in order to hive off basically the categorised rule does the

relative horizontal district of quality to separate and arrange in an order

as the theoretical foundation, observe different district to collect group

through degree , is it hive off the result degree to guarantee at the same

time. And then combine prices to consider and compare and assess syntheti-

cally , try to get the efficiency of every factory brand DVD / price in the

research object Capacity/Price (C/P) ratio. Channel into separately Support

 Vector Machines Regression (SVMR), comprehensive application, as is it

record quality decline analysis of trend , offer user decline to DVD quality

phenomenon make follow-up track to cook, utilize this to follow the trail of

the mechanism , help users to back up early before DVD quality damage, risk

of reducing the materials and losing.

 

沒有留言:

張貼留言