SCMS SCHOOL OF TECHNOLOGY AND MANAGEMENT

SCMS Campus, Prathap Nagar, Muttom, Aluva, COCHIN- !" #$ % &E'ALA

MASTE' OF COMP(TE' APPLICATIONS MCA SEMESTE' )* - SESSION PLAN CO('SE NAME+ DATA MINING CO('SE CODE + MCA 504 FAC(LTY+ J.M.LAKSHMI MAHESH
COURSE OBJECTIVES
The objectives of this course are to: 1) Make the students to be aware of the concept of Data Minin ! Data "are house # $% To adopt &arious techni'ues for c(usterin and )ssociation *% To fa+i(iari,e the app(ication areas of Data Minin in business

#
COURSE CONTENT: Module I - Introduction to Data inin! " Date #arehouse "hat is Data +inin - Data +inin .On "hat kinds of Data- Data +inin /unctiona(ities- C(assification of Data +inin S0ste+s- Data Minin Task 1ri+iti&es- Inte ration of Data +inin s0ste+s- Ma2or issues of Data +inin - "hat is Data "arehouse- Mu(tidi+ensiona( Data Mode(- ) three.tier Data "arehousin )rchitecture .

.
Module II$Data %re&rocessin! and Minin! 're(uent %atterns Data %re&rocessin! Data C(eanin - Data Inte ration and Transfor+ation- Data Reduction- Data discreti,ation and concept hierarch0 eneration# )ssociation Rules Basic Concepts- Efficient and Sca(ab(e /re'uent Ite+ set Minin Methods 3 )priori )( orith+- 4eneratin association Ru(es fro+ /re'uent Ite+ sets- I+pro&in the Efficienc0 of )priori# Minin /re'uent ite+.sets without Candidate 4eneration Module III-Classification and %rediction Introduction to C(assification and 1rediction- Issues Re ardin C(assification and 1rediction Classification b* Decision Tree Induction: Decision Tree induction- )ttribute Se(ection Measures- Tree 1runin - +a*esian Classification: Ba0es5 theore+- 6a7&e Ba0esian C(assification- Rule +ased Algorithms: Usin If . Then ru(es of C(assification- Ru(e E8traction fro+ a Decision Tree- Ru(e Induction Usin a Se'uentia( Co&erin a( orith+- 9. 6earest 6ei hbour C(assifiers# Prediction : :inear Re ression- 6on(inear Re ression- Other Re ression.Based Methods Module I, - Clusterin! #hat is Cluster )nal*sis- Re'uire+ents of C(uster )na(0sis5 T0pes of Data in C(uster )na(0sis- Cate ori,ation of Ma2or C(usterin Methods, %artitionin! Methods :k.Means and k. Medoids- /ro+ 9.Medoids to C:)R)6S , .ierarchical Method : ) (o+erati&e and Di&isi&e ;ierarchica( C(usterin - BIRC;- ROC9- Cha+e(eonDensit*$+ased Method: DBSCAN, /rid +ased Methods: STING: STatistica( I6for+ation 4rid"a&e C(uster, Model based Methods-E8pectation.Ma8i+i,ation- Conceptua( C(usterin - 6eura( 6etwork )pproach# Module ,-)&&lications and Trends in Data Minin! Data Minin! )&&lications : Data Minin for /inancia( Data )na(0sis- Data Minin for the Retai( Industr0- Data Minin for the Te(eco++unication Industr0- Data Minin for Bio(o ica( Data )na(0sis- Data Minin in Other Scientific

)pp(ications- Data Minin for Intrusion Detection- Socia( I+pacts of Data Minin - Trends in Data Minin # Te0t +oo1s and References : 1.Data Minin Concepts and Techni'ues < Jiawei ;an and Miche(ine 9a+ber- Second Edition- E(se&ier- $==> $#Data Minin < B1B Editoria( Board- B1B 1ub(ications- /irst Edition- $==? *#Data "arehousin - Data Minin - ! O:)1 < )(e8 Berson- Stephen J S+ith- Tata Mc4raw ;i((- $==? ?#Data "arehousin - Sinha- Tho+son :earnin - /irst Edn . SESSION)2 )SSESSMENT : TOT)2 M)R3S : 45

S,ss-o.al ass,ssm,.t /-ll 0, 0as,1 o.+

)SSESSMENT DET)I2S: )SSESSMENT 6 : Class &erfor ance Markin Criteria3 Based on c(ass 1articipation- )ttenti&eness ! )ttendance . )SSESSMENT 4 : )ssi!n ents )ssess+ent Re'uire+ents3 .Students are instructed to sub+it Indi&idua( )ssi n+ent for this course .To be sub+itted after the co+p(etion of ?th +odu(e- e8act date wi(( be announced (ater .The c(ass is di&ided into @> roups# .Each roup wi(( be assi ned a particu(ar topic as in the (ist i&en be(ow# .Indi&idua( +e+bers of the roup shou(d se(ect one subtopic fro+ the topic assi ned to hisBher roup and sub+it the assi n+ent for the sa+e# )ssi!n ent (uestion :WRITE SHORT NOTES ON THE GIVEN TOPIC: 2ist of To&ics: @# Decision Trees in c(assification a( orith+# $# 1robabi(it0 for Data Miners *# Statistica( data +inin ?# 6eura( 6etworks ! its app(ication in Data Minin D# Re ression )( orith+s ! its app(ication in Data Minin ># Ba0esian 6etworks ! its app(ication in Data Minin E# 6ai&e Ba0esian 6etworks in Data +inin C# 9.+eans and ;ierarchica( C(usterin # F# )rtificia( Inte((i ence ! Data Minin @=# 4enetic )( orith+s ! its app(ication in Data Minin @@# E&o(utionar0 )( orith+s and Data Minin @$# /u,,0 Minin @*# Business Inte((i ence S0ste+ @?# Bi data )na(0sis @D# "E9) Software @># M)T:)B Software %resentation:

1resentation shou(d be i&en on the )ssi n+ent topic a((oted for the indi&idua(s- after the co+p(etion of Dth +odu(e# Ge8act dates wi(( be announced (ater% )SSESSMENT 7 :class tests : $ c(ass tests wi(( be conducted - each carr0in *= +arks. )SSESSMENT 8 : Mid-Ter "i(( be conducted after the co+p(etion of $=th session# )SSESSMENT 5 : Model-E0a "i(( be conducted at the end of the se+ester.

S,ss-o. Pla.+
Top-5 S,ss-o. 6s7 ',8,r,.5,

Module I

.Data Minin Concepts and Techni'ues < Jiawei ;an and Miche(ine 9a+berSecond Edition- E(se&ier- $==> .Data Minin < B1B Editoria( Board- B1B 1ub(ications- /irst Edition- $==? Data "arehousin - Data Minin - ! O:)1 < )(e8 Berson- Stephen J S+ith- Tata Mc4raw ;i((- $==? .Data "arehousin - Sinha- Tho+son :earnin - /irst Edn. @-$ *-? D >-E C F @=-@@ @$ .Data Minin Concepts and Techni'ues < Jiawei ;an and Miche(ine 9a+berSecond Edition- E(se&ier- $==> .Data Minin < B1B Editoria( Board- B1B 1ub(ications- /irst Edition- $==? @* @? @D @> @E @C-@F $= $@ $$

Introduction to Data Minin and its i+portance in softwareco+puter Industr0 kinds of DataData +inin /unctiona(ities C(assification of Data +inin S0ste+s- Data Minin Task 1ri+iti&es Inte ration of Data +inin s0ste+s- Ma2or issues of Data +inin Data "arehouse Mu(tidi+ensiona( Data Mode(- ) three.tier Data "arehousin )rchitecture. C(ass test.@ Module II

Data 1reprocessin and its need Data C(eanin Data Inte ration and Transfor+ation Data ReductionData discreti,ation and concept hierarch0 eneration# Introduction to )ssociation Ru(es- Efficient and Sca(ab(e /re'uent Ite+ set- Minin Methods )priori )( orith+- I+pro&in the Efficienc0 of )priori# 4eneratin association Ru(es fro+ /re'uent Ite+ sets /re'uent ite+.sets without Candidate 4eneration

Module III

.Data Minin Concepts and Techni'ues < Jiawei ;an and Miche(ine 9a+berSecond Edition- E(se&ier- $==> .Data Minin < B1B Editoria( Board- B1B 1ub(ications- /irst Edition- $==? $* $? $D $>-$E $C $F-*= *@ *$ .Data Minin Concepts and Techni'ues < Jiawei ;an and Miche(ine 9a+berSecond Edition- E(se&ier- $==> .Data Minin < B1B Editoria( Board- B1B 1ub(ications- /irst Edition- $==? ** *?*D *> *E *C *F ?= Data Minin Concepts and Techni'ues < Jiawei ;an and Miche(ine 9a+berSecond Edition- E(se&ier- $==> .Data Minin < B1B Editoria( Board- B1B 1ub(ications- /irst Edition- $==? . Data "arehousin - Data Minin - ! O:)1 < )(e8 Berson- Stephen J S+ithTata Mc4raw ;i((- $==? ?@ ?$ ?* ?? ?D ?>.D=

Introduction to C(assification and 1rediction Issues Re ardin C(assification and 1rediction C(assification b0 Decision Tree Induction Ba0esian C(assification Ru(e Based )( orith+s3 - Ru(e Induction Usin a Se'uentia( Co&erin a( orith+9. 6earest 6ei hbour C(assifiers# 1rediction and its +ode(s Module I

Introduction to c(usters- C(uster ana(0sis-re'uire+ents T0pes of Data in C(uster )na(0sis1artitionin Methods ;ierarchica( Method Densit0<Based Method 4rid Based Methods3 Mode( based Methods C(ass test.$ Module #

Data +inin )pp(ications in &arious fie(ds- /inancia( Data )na(0sis- Data Minin for the Retai( Industr0- Data Minin for the Te(eco++unication Industr0Data Minin for Bio(o ica( Data )na(0sis- Data Minin in Other Scientific )pp(icationsData Minin for Intrusion Detection- Socia( I+pacts of Data Minin Trends in Data Minin # 1resentations