Project Presentation

## Recurrent Neural Network Model<br>for Weather Forecasting<br>Using ISRT Weather Data

.right[
### <br>Md Mutasim Billah
.large[Class Roll: SH-033-030<br>Session: 2022-23<br>Institute of Statistical Research and Training<br>University of Dhaka]]

---

## .center[Overview]

.centermiddle[.laarge[
<svg aria-hidden="true" role="img" viewBox="0 0 640 512" style="height:1em;width:1.25em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:#01161E;overflow:visible;position:relative;"><path d="M294.2 1.2c5.1 2.1 8.7 6.7 9.6 12.1l14.1 84.7 84.7 14.1c5.4 .9 10 4.5 12.1 9.6s1.5 10.9-1.6 15.4l-38.5 55c-2.2-.1-4.4-.2-6.7-.2c-23.3 0-45.1 6.2-64 17.1l0-1.1c0-53-43-96-96-96s-96 43-96 96s43 96 96 96c8.1 0 15.9-1 23.4-2.9c-36.6 18.1-63.3 53.1-69.8 94.9l-24.4 17c-4.5 3.2-10.3 3.8-15.4 1.6s-8.7-6.7-9.6-12.1L98.1 317.9 13.4 303.8c-5.4-.9-10-4.5-12.1-9.6s-1.5-10.9 1.6-15.4L52.5 208 2.9 137.2c-3.2-4.5-3.8-10.3-1.6-15.4s6.7-8.7 12.1-9.6L98.1 98.1l14.1-84.7c.9-5.4 4.5-10 9.6-12.1s10.9-1.5 15.4 1.6L208 52.5 278.8 2.9c4.5-3.2 10.3-3.8 15.4-1.6zM144 208a64 64 0 1 1 128 0 64 64 0 1 1 -128 0zM639.9 431.9c0 44.2-35.8 80-80 80H288c-53 0-96-43-96-96c0-47.6 34.6-87 80-94.6l0-1.3c0-53 43-96 96-96c34.9 0 65.4 18.6 82.2 46.4c13-9.1 28.8-14.4 45.8-14.4c44.2 0 80 35.8 80 80c0 5.9-.6 11.7-1.9 17.2c37.4 6.7 65.8 39.4 65.8 78.7z"/></svg> **What we wanted to do**

---
.center[
## Weather Forecast
]

.pull-left[
.center[.xlarge[<svg aria-hidden="true" role="img" viewBox="0 0 320 512" style="height:2em;width:1.25em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:currentColor;overflow:visible;position:relative;"><path d="M160 64c-26.5 0-48 21.5-48 48V276.5c0 17.3-7.1 31.9-15.3 42.5C86.2 332.6 80 349.5 80 368c0 44.2 35.8 80 80 80s80-35.8 80-80c0-18.5-6.2-35.4-16.7-48.9c-8.2-10.6-15.3-25.2-15.3-42.5V112c0-26.5-21.5-48-48-48zM48 112C48 50.2 98.1 0 160 0s112 50.1 112 112V276.5c0 .1 .1 .3 .2 .6c.2 .6 .8 1.6 1.7 2.8c18.9 24.4 30.1 55 30.1 88.1c0 79.5-64.5 144-144 144S16 447.5 16 368c0-33.2 11.2-63.8 30.1-88.1c.9-1.2 1.5-2.2 1.7-2.8c.1-.3 .2-.5 .2-.6V112zM208 368c0 26.5-21.5 48-48 48s-48-21.5-48-48c0-20.9 13.4-38.7 32-45.3V208c0-8.8 7.2-16 16-16s16 7.2 16 16V322.7c18.6 6.6 32 24.4 32 45.3z"/></svg><br>Temperature<br><br><svg aria-hidden="true" role="img" viewBox="0 0 384 512" style="height:2em;width:1.5em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:currentColor;overflow:visible;position:relative;"><path d="M192 512C86 512 0 426 0 320C0 228.8 130.2 57.7 166.6 11.7C172.6 4.2 181.5 0 191.1 0h1.8c9.6 0 18.5 4.2 24.5 11.7C253.8 57.7 384 228.8 384 320c0 106-86 192-192 192zM96 336c0-8.8-7.2-16-16-16s-16 7.2-16 16c0 61.9 50.1 112 112 112c8.8 0 16-7.2 16-16s-7.2-16-16-16c-44.2 0-80-35.8-80-80z"/></svg><br>Humidity<br><br><svg aria-hidden="true" role="img" viewBox="0 0 448 512" style="height:2em;width:1.75em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:currentColor;overflow:visible;position:relative;"><path d="M176 0c-17.7 0-32 14.3-32 32s14.3 32 32 32h16V98.4C92.3 113.8 16 200 16 304c0 114.9 93.1 208 208 208s208-93.1 208-208c0-41.8-12.3-80.7-33.5-113.2l24.1-24.1c12.5-12.5 12.5-32.8 0-45.3s-32.8-12.5-45.3 0L355.7 143c-28.1-23-62.2-38.8-99.7-44.6V64h16c17.7 0 32-14.3 32-32s-14.3-32-32-32H224 176zm72 192V320c0 13.3-10.7 24-24 24s-24-10.7-24-24V192c0-13.3 10.7-24 24-24s24 10.7 24 24z"/></svg><br>Pressure
]
]
]

.pull-right[
.center[.xlarge[<svg aria-hidden="true" role="img" viewBox="0 0 512 512" style="height:2em;width:2em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:currentColor;overflow:visible;position:relative;"><path d="M96 320c-53 0-96-43-96-96c0-42.5 27.6-78.6 65.9-91.2C64.7 126.1 64 119.1 64 112C64 50.1 114.1 0 176 0c43.1 0 80.5 24.3 99.2 60c14.7-17.1 36.5-28 60.8-28c44.2 0 80 35.8 80 80c0 5.5-.6 10.8-1.6 16c.5 0 1.1 0 1.6 0c53 0 96 43 96 96s-43 96-96 96H96zm-6.8 52c1.3-2.5 3.9-4 6.8-4s5.4 1.5 6.8 4l35.1 64.6c4.1 7.5 6.2 15.8 6.2 24.3v3c0 26.5-21.5 48-48 48s-48-21.5-48-48v-3c0-8.5 2.1-16.9 6.2-24.3L89.2 372zm160 0c1.3-2.5 3.9-4 6.8-4s5.4 1.5 6.8 4l35.1 64.6c4.1 7.5 6.2 15.8 6.2 24.3v3c0 26.5-21.5 48-48 48s-48-21.5-48-48v-3c0-8.5 2.1-16.9 6.2-24.3L249.2 372zm124.9 64.6L409.2 372c1.3-2.5 3.9-4 6.8-4s5.4 1.5 6.8 4l35.1 64.6c4.1 7.5 6.2 15.8 6.2 24.3v3c0 26.5-21.5 48-48 48s-48-21.5-48-48v-3c0-8.5 2.1-16.9 6.2-24.3z"/></svg><br>Rain<br><br><svg aria-hidden="true" role="img" viewBox="0 0 512 512" style="height:2em;width:2em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:currentColor;overflow:visible;position:relative;"><path d="M288 32c0 17.7 14.3 32 32 32h32c17.7 0 32 14.3 32 32s-14.3 32-32 32H32c-17.7 0-32 14.3-32 32s14.3 32 32 32H352c53 0 96-43 96-96s-43-96-96-96H320c-17.7 0-32 14.3-32 32zm64 352c0 17.7 14.3 32 32 32h32c53 0 96-43 96-96s-43-96-96-96H32c-17.7 0-32 14.3-32 32s14.3 32 32 32H416c17.7 0 32 14.3 32 32s-14.3 32-32 32H384c-17.7 0-32 14.3-32 32zM128 512h32c53 0 96-43 96-96s-43-96-96-96H32c-17.7 0-32 14.3-32 32s14.3 32 32 32H160c17.7 0 32 14.3 32 32s-14.3 32-32 32H128c-17.7 0-32 14.3-32 32s14.3 32 32 32z"/></svg><br>Wind<br><br><svg aria-hidden="true" role="img" viewBox="0 0 640 512" style="height:2em;width:2.5em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:currentColor;overflow:visible;position:relative;"><path d="M0 336c0 79.5 64.5 144 144 144H512c70.7 0 128-57.3 128-128c0-61.9-44-113.6-102.4-125.4c4.1-10.7 6.4-22.4 6.4-34.6c0-53-43-96-96-96c-19.7 0-38.1 6-53.3 16.2C367 64.2 315.3 32 256 32C167.6 32 96 103.6 96 192c0 2.7 .1 5.4 .2 8.1C40.2 219.8 0 273.2 0 336z"/></svg><br>Cloud(!)
]]]

.footnote[<a name=cite-ncas_weather_components></a>[for
Atmospheric Science (2023)](#bib-ncas_weather_components)]

---

## .center[Literature Review]

### .box[.blue[.center[
Neural Network Model<br>over Classical Statistical Methods
]]]

.box[
.large[.center[<a name=cite-hanoon_developing_2021></a>[Hanoon, Ahmed, Zaini, Razzaq, Kumar, Sherif, Sefelnasr, and El-Shafie (2021)](https://doi.org/10.1038/s41598-021-96872-w); <a name=cite-intan_performance_2021></a>[Intan, Rismayani, Ghani, Nurdin, and Koswara (2021)](https://www.neliti.com/publications/519180/); <a name=cite-kakar_artificial_2018></a>[Kakar, Sheikh, Naseem, Iqbal, Rehman, Kakar, Kakar, Kakar, and Khan (2018)](https://thesai.org/Publications/ViewPaper?Volume=9&Issue=8&Code=IJACSA&SerialNo=59); <a name=cite-hayati_application_2007></a>[Hayati and Mohebi (2007)](#bib-hayati_application_2007); <a name=cite-roy_forecasting_2020></a>[Roy (2020)](https://doi.org/10.1016/j.procs.2020.11.005); <a name=cite-nketiah_recurrent_2023></a>[Nketiah, Chenlong, Yingchuan, and Aram (2023)](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0285713);<a name=cite-zaytar_sequence_2016></a>[Zaytar and El Amrani (2016)](https://doi.org/10.5120/ijca2016910497)
]]]

---

## .center[
Our Goal]

.centermiddle[.laarge[
<svg aria-hidden="true" role="img" viewBox="0 0 512 512" style="height:1em;width:1em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:#01161E;overflow:visible;position:relative;"><path d="M448 256A192 192 0 1 0 64 256a192 192 0 1 0 384 0zM0 256a256 256 0 1 1 512 0A256 256 0 1 1 0 256zm256 80a80 80 0 1 0 0-160 80 80 0 1 0 0 160zm0-224a144 144 0 1 1 0 288 144 144 0 1 1 0-288zM224 256a32 32 0 1 1 64 0 32 32 0 1 1 -64 0z"/></svg> Build RNN model<br>for several weather components

]]

---

## .center[Overview]

.centermiddle[.laarge[
.fade[<svg aria-hidden="true" role="img" viewBox="0 0 640 512" style="height:1em;width:1.25em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:#01161E;overflow:visible;position:relative;"><path d="M294.2 1.2c5.1 2.1 8.7 6.7 9.6 12.1l14.1 84.7 84.7 14.1c5.4 .9 10 4.5 12.1 9.6s1.5 10.9-1.6 15.4l-38.5 55c-2.2-.1-4.4-.2-6.7-.2c-23.3 0-45.1 6.2-64 17.1l0-1.1c0-53-43-96-96-96s-96 43-96 96s43 96 96 96c8.1 0 15.9-1 23.4-2.9c-36.6 18.1-63.3 53.1-69.8 94.9l-24.4 17c-4.5 3.2-10.3 3.8-15.4 1.6s-8.7-6.7-9.6-12.1L98.1 317.9 13.4 303.8c-5.4-.9-10-4.5-12.1-9.6s-1.5-10.9 1.6-15.4L52.5 208 2.9 137.2c-3.2-4.5-3.8-10.3-1.6-15.4s6.7-8.7 12.1-9.6L98.1 98.1l14.1-84.7c.9-5.4 4.5-10 9.6-12.1s10.9-1.5 15.4 1.6L208 52.5 278.8 2.9c4.5-3.2 10.3-3.8 15.4-1.6zM144 208a64 64 0 1 1 128 0 64 64 0 1 1 -128 0zM639.9 431.9c0 44.2-35.8 80-80 80H288c-53 0-96-43-96-96c0-47.6 34.6-87 80-94.6l0-1.3c0-53 43-96 96-96c34.9 0 65.4 18.6 82.2 46.4c13-9.1 28.8-14.4 45.8-14.4c44.2 0 80 35.8 80 80c0 5.9-.6 11.7-1.9 17.2c37.4 6.7 65.8 39.4 65.8 78.7z"/></svg> What we wanted to do]

---

## Data

.panel[.panel-name[Feature Engineering]
.pull-left[.center[.laarge[
T2M<br><br>
PRECTOTCORR<br><br>
Wx<br><br>
Day sin<br><br>
Year sin
]]]

QV2M<br><br> 
PS<br><br>
Wy<br><br>
Day cos<br><br>
Year cos
]]]
]

<img src="data:image/png;base64,#img/f-4.png" width="70%" />
.center[.xlarge[Features after normalization]]
]

.panel[.panel-name[Data Split]
.laarge[.box[.blue[.center[
Training Set <svg aria-hidden="true" role="img" viewBox="0 0 512 512" style="height:1em;width:1em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:#01161E;overflow:visible;position:relative;"><path d="M502.6 278.6c12.5-12.5 12.5-32.8 0-45.3l-128-128c-12.5-12.5-32.8-12.5-45.3 0s-12.5 32.8 0 45.3L402.7 224 32 224c-17.7 0-32 14.3-32 32s14.3 32 32 32l370.7 0-73.4 73.4c-12.5 12.5-12.5 32.8 0 45.3s32.8 12.5 45.3 0l128-128z"/></svg> 70%
]]]]
.laarge[.box[.blue[.center[
Validation Set <svg aria-hidden="true" role="img" viewBox="0 0 512 512" style="height:1em;width:1em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:#01161E;overflow:visible;position:relative;"><path d="M502.6 278.6c12.5-12.5 12.5-32.8 0-45.3l-128-128c-12.5-12.5-32.8-12.5-45.3 0s-12.5 32.8 0 45.3L402.7 224 32 224c-17.7 0-32 14.3-32 32s14.3 32 32 32l370.7 0-73.4 73.4c-12.5 12.5-12.5 32.8 0 45.3s32.8 12.5 45.3 0l128-128z"/></svg> 20%
]]]]
.laarge[.box[.blue[.center[
Test Set <svg aria-hidden="true" role="img" viewBox="0 0 512 512" style="height:1em;width:1em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:#01161E;overflow:visible;position:relative;"><path d="M502.6 278.6c12.5-12.5 12.5-32.8 0-45.3l-128-128c-12.5-12.5-32.8-12.5-45.3 0s-12.5 32.8 0 45.3L402.7 224 32 224c-17.7 0-32 14.3-32 32s14.3 32 32 32l370.7 0-73.4 73.4c-12.5 12.5-12.5 32.8 0 45.3s32.8 12.5 45.3 0l128-128z"/></svg> 10%
]]]]
]

]

---
class: center

## Model

.panelset[
.panel[.panel-name[LSTM]
<img src="data:image/png;base64,#img/multistep_last.png" width="50%" />
.center[.xlarge[LSTM Model]]
]
.panel[.panel-name[AR LSTM]
<img src="data:image/png;base64,#img/multistep_autoregressive.png" width="70%" />
.center[.xlarge[Autoregressive LSTM Model]]
]

]

---

## .center[Overview]

---

## Model Evaluation

----

### **Mean Squared Error**

.xlarge[LSTM <svg aria-hidden="true" role="img" viewBox="0 0 512 512" style="height:1em;width:1em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:#01161E;overflow:visible;position:relative;"><path d="M502.6 278.6c12.5-12.5 12.5-32.8 0-45.3l-128-128c-12.5-12.5-32.8-12.5-45.3 0s-12.5 32.8 0 45.3L402.7 224 32 224c-17.7 0-32 14.3-32 32s14.3 32 32 32l370.7 0-73.4 73.4c-12.5 12.5-12.5 32.8 0 45.3s32.8 12.5 45.3 0l128-128z"/></svg> 0.0099<br>AR LSTM <svg aria-hidden="true" role="img" viewBox="0 0 512 512" style="height:1em;width:1em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:#01161E;overflow:visible;position:relative;"><path d="M502.6 278.6c12.5-12.5 12.5-32.8 0-45.3l-128-128c-12.5-12.5-32.8-12.5-45.3 0s-12.5 32.8 0 45.3L402.7 224 32 224c-17.7 0-32 14.3-32 32s14.3 32 32 32l370.7 0-73.4 73.4c-12.5 12.5-12.5 32.8 0 45.3s32.8 12.5 45.3 0l128-128z"/></svg> 0.0093]
]

---

## Model Evaluation

----

### **Mean Absolute Error**

.xlarge[LSTM <svg aria-hidden="true" role="img" viewBox="0 0 512 512" style="height:1em;width:1em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:#01161E;overflow:visible;position:relative;"><path d="M502.6 278.6c12.5-12.5 12.5-32.8 0-45.3l-128-128c-12.5-12.5-32.8-12.5-45.3 0s-12.5 32.8 0 45.3L402.7 224 32 224c-17.7 0-32 14.3-32 32s14.3 32 32 32l370.7 0-73.4 73.4c-12.5 12.5-12.5 32.8 0 45.3s32.8 12.5 45.3 0l128-128z"/></svg> 0.0586<br>AR LSTM <svg aria-hidden="true" role="img" viewBox="0 0 512 512" style="height:1em;width:1em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:#01161E;overflow:visible;position:relative;"><path d="M502.6 278.6c12.5-12.5 12.5-32.8 0-45.3l-128-128c-12.5-12.5-32.8-12.5-45.3 0s-12.5 32.8 0 45.3L402.7 224 32 224c-17.7 0-32 14.3-32 32s14.3 32 32 32l370.7 0-73.4 73.4c-12.5 12.5-12.5 32.8 0 45.3s32.8 12.5 45.3 0l128-128z"/></svg> 0.0558]
]

---

## Concluding Remark

---
## .center[Reference-I]
.large[
<a name=bib-ncas_weather_components></a>[Atmospheric Science, T. N. C.
for](#cite-ncas_weather_components) (2023). _What causes weather?_.
<https://ncas.ac.uk/learn/what-causes-weather/>. [Online; accessed
03-February-2024].

<a name=bib-hanoon_developing_2021></a>[Hanoon, M. S., A. N. Ahmed, N.
Zaini, et al.](#cite-hanoon_developing_2021) (2021). "Developing
machine learning algorithms for meteorological temperature and humidity
forecasting at Terengganu state in Malaysia". Eng. In: _Scientific
Reports_ 11.1, p. 18935. ISSN: 2045-2322. DOI:
[10.1038/s41598-021-96872-w](https://doi.org/10.1038%2Fs41598-021-96872-w).

<a name=bib-hayati_application_2007></a>[Hayati, M. and Z.
Mohebi](#cite-hayati_application_2007) (2007). "Application of
Artificial Neural Networks for Temperature Forecasting". In: _World
Academy of Science, Engineering and Technology_ 28, pp. 275-279.
]

---
## .center[Reference-II]
.large[
<a name=bib-intan_performance_2021></a>[Intan, I., R. Rismayani, S. A.
D. Ghani, et al.](#cite-intan_performance_2021) (2021). "Performance
Analysis of Weather Forecasting Using Machine Learning Algorithms
(Analisis Performansi Prakiraan Cuaca Menggunakan Algoritma Machine
Learning)". En. In: _Pekommas_ 6.2. Publisher: Indonesian Ministry of
Communication and Informatics, pp. 1-8. ISSN: 2502-1893, 2502-1907.
DOI:
[10.30818/jpkm.2021.2060221](https://doi.org/10.30818%2Fjpkm.2021.2060221).
URL:
[https://www.neliti.com/publications/519180/](https://www.neliti.com/publications/519180/)
(visited on Jan. 21, 2024).

<a name=bib-kakar_artificial_2018></a>[Kakar, S. A., N. Sheikh, A.
Naseem, et al.](#cite-kakar_artificial_2018) (2018). "Artificial Neural
Network based Weather Prediction using Back Propagation Technique". En.
In: _International Journal of Advanced Computer Science and
Applications (IJACSA)_ 9.8. Number: 8 Publisher: The Science and
Information (SAI) Organization Limited. ISSN: 2156-5570. DOI:
[10.14569/IJACSA.2018.090859](https://doi.org/10.14569%2FIJACSA.2018.090859).
URL:
[https://thesai.org/Publications/ViewPaper?Volume=9&Issue=8&Code=IJACSA&SerialNo=59](https://thesai.org/Publications/ViewPaper?Volume=9&Issue=8&Code=IJACSA&SerialNo=59)
(visited on Jan. 21, 2024).
]

---
## .center[Reference-III]
.large[
<a name=bib-nketiah_recurrent_2023></a>[Nketiah, E. A., L. Chenlong, J.
Yingchuan, et al.](#cite-nketiah_recurrent_2023) (2023). "Recurrent
neural network modeling of multivariate time series and its application
in temperature forecasting". En. In: _PLOS ONE_ 18.5. Publisher: Public
Library of Science, p. e0285713. ISSN: 1932-6203. DOI:
[10.1371/journal.pone.0285713](https://doi.org/10.1371%2Fjournal.pone.0285713).
URL:
[https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0285713](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0285713)
(visited on Jan. 13, 2024).

<a name=bib-roy_forecasting_2020></a>[Roy,
D.](#cite-roy_forecasting_2020) (2020). "Forecasting The Air
Temperature at a Weather Station Using Deep Neural Networks". In:
_Procedia Computer Science_ 178, pp. 38-46. DOI:
[10.1016/j.procs.2020.11.005](https://doi.org/10.1016%2Fj.procs.2020.11.005).

<a name=bib-zaytar_sequence_2016></a>[Zaytar, M. A. and C. El
Amrani](#cite-zaytar_sequence_2016) (2016). "Sequence to Sequence
Weather Forecasting with Long Short-Term Memory Recurrent Neural
Networks". In: _International Journal of Computer Applications_ 143,
pp. 7-11. DOI:
[10.5120/ijca2016910497](https://doi.org/10.5120%2Fijca2016910497).
]

---

# Thank YOU!!