First Section
-
17. Kernal | Noise | Threshold (102:58)
-
24.Model in ML | confidence score | Feed data | LBPH algorithm | Converting Image | Uses of function and module | Diffrence in neural networks & LBPH algorithm | Croping Image (96:54)
-
25. Regularizarion techniques | Spliting Data | Overfitting | Validation Cycle | Tuning | EPOCH (130:52)
-
26.Overfitting | About dataset | Pandas | Validation | Cross-Validation | train_test_split Method | list of layers | Back and Forward Propogation | SGD Optimizer | Batch Size (95:06)
-
27.ML - About Fit line | About model | Underfitting and overfitting | Explaining Bias and variance | High Bias | Low variance | Penalizing | L1 and L2 Regularization (112:57)
CNN
-
28.CNN introduction | Model-data relationship | Neural network - ANN (135:17)
-
29.Flatenning | Use cases of CNN | Pooling and its methods | Feature learning | Convolution (104:35)
-
30.Features in CNN | Convolve layer | Kernel unit | Strides | Convolve2d() in scipy.signal() | Convolution | Pooling layer | Flattening layer | Feature extraction in CNN (82:59)
-
31.Functions of (Pooling layer, Convolution layer, Flattening layer) | CNN architecture | Significance of strides | Kernel & filters | Explain Pooling (88:00)
-
32.About convolve layer | uses of kernel | Train the model in CNN | Load Dataset | Data generation | Function of Keras (80:36)
-
34.Multiple layer CNN (like many convolve ,polling etc. , Layer in one data set) | LeNet and AlexNet | Relu function work. (84:10)
-
35.Library is much better than opencv | Used instead of cv2, keras | VGG model | type of VGG model | Dataset is used to train VGG model | vgg16 model | Use of preprocess_input and from which module it is imported | Use of decode_prediction (79:15)
-
36.Resnet50 network | difference between VGG16 and Resnet50 | Create weights with Resnet50 model | Inception v3 network (77:10)
-
37.CRP layer | Need of feature extraction | Idea behind transfer learning | List of all pre-created networks | Command for getting the summary (in tabular format) | Layer in keras is trained or not | Lock all the pre-trained layers | Top model layer
-
38.Transfer Learning | ImageNet | Significance of architecture of neural networks in Transfer Learning | Model | Epoch (83:31)
-
39.Logs | DOS Attack | Unsupervised learning | Clustering | Explain the visual | Finding the number of groups (87:18)
-
40.Unsupervised learning | K means clustering | Difference between Cluster and groups | Trained the model to predict satisfaction and loyalty (93:51)
-
New Lecture
RNN
-
41. (86:12)
-
42.Deep Learning Networks | RNN | SequenceSeries dataset | Time Series dataset (84:46)
-
43.LSTM Algorithm | Predict y in RNN | Minimum Requirement of Dataset in RNN for Prediction | Window Period | Logic Behind RNN (87:05)
-
44.Describe RNN and LSTM | How LSTM works | Use of return_sequences=True | Create the model of LSTM | Activation function used in LSTM (60:59)
-
45.SVM - Works and Advantage | Maximum margin | Support vector | Specify use of SVM and SVR | EDA method (82:51)
-
46.Dimension reduction | PCA | Type of logic PCA use behind the scene for Dimension reduction | Import PCA and call it in our project (70:09)
-
47. About PCA | Challenges in PCA | Use of LDA | Modellearning PCA and LDA are based | Use of Gradient Boosting Algorithms | XGBpost Algorithm (64:25)
-
48. Cross validation uses to get better accuracy | Types of cross validation | Process of k-fold cross validation | Library to be imported to increase the folds | Demerits of confusion matrix_DataScience - 12-4-2022 (59:58)
-
49. Difference between Descriptive DS and Inferential Statistics | Hierarchical Clustering | Dendrogram | Dendrogram functionality_DataScience - 13-4-2022 (58:13)
-
50. DataScience - 20-4-2022 (61:57)
Extra Sessions
-
Neural network | Data visualization | Graph | Folium | Object detection | CNN | YOLO-Setup and Use | Optimizer | Initializer (257:08)
-
Data visualization | Classification | Histogram | Graph | Heatmap | Regression | Binary classification (265:33)
-
Binary Classification, Model using sigmoid function | Logistic regression | Visuals | Searborn | Graph | Bar graph | Discrete variable | Frequency distribution table | One hot encoding | Feature Engineering | Missing value resolving using feature engineering | Imputation (123:46)
-
Grafs & Visuals - Part-4 (118:53)
-
Graph | Folium | Seaborn | Pandas | Plotly | Cuffling (115:46)
-
Grafs & Visuals - Part-6 (130:02)
-
Dimensity reduction | Feature selection | Feature extraction | Linear regression | P-value and significance level | OLS and backward elimination and wrapper methond (163:54)
-
Advance CNN - Mask-R-CNN | Feature Learning | WCSS (126:04)
-
GANS, CNN (129:34)
-
GAN (138:56)
-
Association rules | Apyori | Lift (119:21)
-
RBM | CF | Context Aware | Content Based | Association rules | Recommendation system | Matrix factorization | NN (171:05)
-
Neurl Network | RBM | Linear function (121:57)
-
Mask - RCNN (171:58)
-
Random Forest (38:59)
NLP Training
