Merge pull request #557 from CoderOMaster/main

GOOGLE DEVICES Q AND A ANALYSIS
abhisheks008 · Jan 28, 2024 · 8c86a5d · 8c86a5d
2 parents ee6628d + 074e258
commit 8c86a5d
Show file tree

Hide file tree

Showing 8 changed files with 4,314 additions and 0 deletions.
diff --git a/Google Devices Q and A Analysis/Dataset/All About Google Devices - Sheet1.csv b/Google Devices Q and A Analysis/Dataset/All About Google Devices - Sheet1.csv
diff --git a/Google Devices Q and A Analysis/Images/1.png b/Google Devices Q and A Analysis/Images/1.png
diff --git a/Google Devices Q and A Analysis/Images/2.png b/Google Devices Q and A Analysis/Images/2.png
diff --git a/Google Devices Q and A Analysis/Images/3.png b/Google Devices Q and A Analysis/Images/3.png
diff --git a/Google Devices Q and A Analysis/Images/4.png b/Google Devices Q and A Analysis/Images/4.png
diff --git a/Google Devices Q and A Analysis/Model/GoogleQ_Aipynb.ipynb b/Google Devices Q and A Analysis/Model/GoogleQ_Aipynb.ipynb
diff --git a/Google Devices Q and A Analysis/README.md b/Google Devices Q and A Analysis/README.md
@@ -0,0 +1,84 @@
+**GOOGLE Q AND A ANALYSIS**
+
+
+**GOAL**
+
+To build nlp based ml models for google question and answer dataset to specify category of device it predicts.
+
+**DATASET**
+
+https://www.kaggle.com/datasets/aashidutt3/google-devices-q-and-a-dataset
+
+**DESCRIPTION**
+
+To analyze question asked by customer along with device to predict which category problem falls onto wrt google dataset.
+
+**Visualization and EDA of different attributes:**
+
+
+![Alt Text](./Images/1.png)
+
+![Alt Text](./Images/2.png)
+
+![Alt Text](./Images/3.png)
+
+![Alt Text](./Images/4.png)
+
+
+
+**MODELS USED**
+1) Decision Tree 
+2) Random Forest
+3) Catboost
+4) Multivariate Logisitic Regression
+5) Gradient Boost
+6) Ridge
+7) MLP
+8) XGBoost
+
+**LIBRARIES**
+1) PANDAS
+2) NUMPY
+3) CATBOOST
+4) MATPLOTLIB
+5) OS
+6) SCIKIT-LEARN
+7) XGBOOST
+8) SPACY
+
+**IMPLEMENTATION**
+1) Loaded the dataset with 3 columns having 997 entries.
+2) Implementated EDA with visualisation using wordcloud to get estimate of important labels to be used later on for classification.
+3) Duplicate values were taken care of and text cleaning was implemented.
+4) Distortion score elbow for K-means was implemented.
+5) Tokenization and vectorization of textual data was implemented
+6) Model was trained efficiently with different machine learning algorithms
+
+
+**RESULT**
+1) Multivariate Logisitic Regression calculated exceptional accuracy of 0.86
+2) Rest of the models performed fairly well wrt to smaller database.
+
+
+
+...
+
+### Models and Accuracies
+
+| Model | Accuracy | 
+| ----------------------------- |:----------:|
+| Decision Tree | 0.76 | 
+| Random Forest | 0.7592 | 
+| Logistic Regression(Multi) | 0.86 | 
+| Gradient Boost | 0.796 | 
+| Ridge | 0.82 | 
+| MLP | 0.83 | 
+| XGBoost | 0.798 |
+| Catboost | 0.85 |
+
+
+...
+
+**NAME**
+
+Keshav Arora
diff --git a/Google Devices Q and A Analysis/requirements.txt b/Google Devices Q and A Analysis/requirements.txt
@@ -0,0 +1,4 @@
+pandas==1.3.3
+matplotlib==3.4.3
+numpy==1.21.2
+catboost==1.2.2