?url_ver=Z39.88-2004&rft_id=2117031063&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&rft.title=TIME+SERIES+ANALYSIS+USING+CONVOLUTIONAL+NEURAL%0D%0ANETWORKS+(CNN)+AND+RECURRENT+NEURAL+NETWORKS+(RNN)%0D%0AFOR+MODELING+THE+FREQUENCY+OF+INFECTIOUS+DISEASE%0D%0AEPIDEMIC+NEWS&rft.creator=SEPTIA+DAMAYANTI%2C+CLARISA+&rft.subject=500+ilmu+pengetahuan+alam+dan+matematika&rft.subject=505+Terbitan+berseri+di+bidang+ilmu+pengetahuan+alam&rft.subject=510+Matematika&rft.description=The+COVID-19+pandemic+has+highlighted+the+importance+of+leveraging+online+data%0D%0Aas+a+tool+for+predicting+future+infectious+disease+trends.+This+study+aims+to+compare%0D%0Athe+performance+of+two+deep+learning+methods%2C+namely+Convolutional+Neural%0D%0ANetwork+(CNN)+and+Recurrent+Neural+Network+(RNN)%2C+in+predicting+the+daily+frequency%0D%0Aof+online+news+publications+based+on+three+sentiment+classes%3A+negative%2C+neutral%2C%0D%0Aand+positive.+The+results+show+that+CNN+delivers+the+best+performance%2C+with%0D%0Aan+RMSE+of+0.14+and+MAPE+of+27%25%2C+demonstrating+its+superiority+in+recognizing%0D%0Acomplex+patterns+in+large+datasets%2C+especially+for+negative+and+neutral+sentiment%0D%0Adata.+Meanwhile%2C+RNN+also+yields+reasonably+good+results%2C+particularly+for+smaller%0D%0Adatasets+such+as+those+with+positive+sentiment%2C+although+with+slightly+lower+accuracy%0D%0A(RMSE+of+0.17+and+MAPE+of+35%25).+These+findings+suggest+that+CNN+is%0D%0Ahighly+recommended+for+predictions+on+large-scale+datasets%2C+while+RNN+serves+as%0D%0Aa+relevant+alternative+when+data+availability+is+limited%2C+albeit+with+a+slightly+lower%0D%0Aaccuracy+rate.+Overall%2C+deep+learning+models+have+proven+effective+in+predicting%0D%0Athe+frequency+of+online+news+publications+based+on+sentiment%2C+supporting+the+use%0D%0Aof+online+news+as+an+alternative+data+source+for+monitoring+public+health+issues.%0D%0AKeywords%3A+COVID-19%2C+infectious+diseases%2C+online+news%2C+sentiment.%0D%0A%0D%0APandemi+COVID-19+telah+menyoroti+pentingnya+pemanfaatan+data+online+sebagai%0D%0Aalat+untuk+memprediksi+tren+penyakit+menular+di+masa+depan.+Studi+ini+bertujuan%0D%0Auntuk+membandingkan+kinerja+dua+metode+deep+learning%2C+yaitu+Convolutional+Neural%0D%0ANetwork+(CNN)+dan+Recurrent+Neural+Network+(RNN)%2C+dalam+memprediksi%0D%0Afrekuensi+harian+publikasi+berita+online+berdasarkan+tiga+kelas+sentimen%3A+negatif%2C%0D%0Anetral%2C+dan+positif.+Hasil+penelitian+menunjukkan+bahwa+CNN+memberikan+kinerja%0D%0Aterbaik%2C+dengan+nilai+RMSE+sebesar+0%2C14+dan+MAPE+sebesar+27%25%2C+menunjukkan%0D%0Akeunggulannya+dalam+mengenali+pola+kompleks+pada+dataset+besar%2C+terutama+untuk%0D%0Adata+sentimen+negatif+dan+netral.+Sementara+itu%2C+RNN+juga+menghasilkan+performa%0D%0Ayang+cukup+baik%2C+khususnya+untuk+dataset+yang+lebih+kecil+seperti+data+dengan+sentimen%0D%0Apositif%2C+meskipun+dengan+tingkat+akurasi+yang+sedikit+lebih+rendah+(RMSE%0D%0Asebesar+0%2C17+dan+MAPE+sebesar+35%25).+Temuan+ini+menunjukkan+bahwa+CNN+sangat%0D%0Adirekomendasikan+untuk+prediksi+pada+dataset+berskala+besar%2C+sementara+RNN%0D%0Amerupakan+alternatif+yang+relevan+ketika+ketersediaan+data+terbatas%2C+meskipun+dengan%0D%0Atingkat+akurasi+yang+sedikit+lebih+rendah.+Secara+keseluruhan%2C+model+deep%0D%0Alearning+terbukti+efektif+dalam+memprediksi+frekuensi+publikasi+berita+online+berdasarkan%0D%0Asentimen%2C+sehingga+mendukung+penggunaan+berita+online+sebagai+sumber%0D%0Adata+alternatif+untuk+pemantauan+isu+kesehatan+masyarakat.%0D%0AKata+kunci%3A+COVID-19%2C+penyakit+menular%2C+berita+online%2C+sentimen.&rft.publisher=FAKULTAS+MATEMATIKA+DAN+ILMU+PENGETAHUAN+ALAM&rft.date=2025-05-20&rft.type=Skripsi&rft.type=NonPeerReviewed&rft.format=text&rft.identifier=http%3A%2F%2Fdigilib.unila.ac.id%2F88535%2F1%2FABSTRAK.pdf&rft.format=text&rft.identifier=http%3A%2F%2Fdigilib.unila.ac.id%2F88535%2F2%2FSKRIPSI%2520FULL.pdf&rft.format=text&rft.identifier=http%3A%2F%2Fdigilib.unila.ac.id%2F88535%2F3%2FSKRIPSI%2520FULL%2520TANPA%2520BAB%2520PEMBAHASAN.pdf&rft.identifier=++SEPTIA+DAMAYANTI%2C+CLARISA+++(2025)+TIME+SERIES+ANALYSIS+USING+CONVOLUTIONAL+NEURAL+NETWORKS+(CNN)+AND+RECURRENT+NEURAL+NETWORKS+(RNN)+FOR+MODELING+THE+FREQUENCY+OF+INFECTIOUS+DISEASE+EPIDEMIC+NEWS.++FAKULTAS+MATEMATIKA+DAN+ILMU+PENGETAHUAN+ALAM%2C+UNIVERSITAS+LAMPUNG.+++++&rft.relation=http%3A%2F%2Fdigilib.unila.ac.id%2F88535%2F