From b19a4c650b76ca4138a0e06d275a289a16104a05 Mon Sep 17 00:00:00 2001 From: Lorenzo Rossi Date: Sat, 19 Oct 2024 16:21:23 +0100 Subject: [PATCH 1/4] Removed specification of version for Python packages in environment.yml --- problem-sets/environment.yml | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/problem-sets/environment.yml b/problem-sets/environment.yml index eed253b..7d75146 100644 --- a/problem-sets/environment.yml +++ b/problem-sets/environment.yml @@ -2,8 +2,8 @@ name: cs229 channels: - defaults dependencies: - - matplotlib=2.2.2 - - numpy=1.15.0 - - pip=10.0.1 - - python=3.6.6 - - scipy=1.1.0 + - matplotlib + - numpy + - pip + - python + - scipy From a8b1a352ad2a898d5466b41ded0584bd9155f0c7 Mon Sep 17 00:00:00 2001 From: Lorenzo Rossi Date: Tue, 5 Nov 2024 18:02:35 +0000 Subject: [PATCH 2/4] In problem-sets-solutions/PS1/code/src/p03d_poisson.py, intercept added to input features x for training and validation set. --- problem-sets-solutions/PS1/code/src/p03d_poisson.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/problem-sets-solutions/PS1/code/src/p03d_poisson.py b/problem-sets-solutions/PS1/code/src/p03d_poisson.py index 83c4b98..b40abaa 100644 --- a/problem-sets-solutions/PS1/code/src/p03d_poisson.py +++ b/problem-sets-solutions/PS1/code/src/p03d_poisson.py @@ -15,14 +15,14 @@ def main(lr, train_path, eval_path, pred_path): pred_path: Path to save predictions. """ # Load training set - x_train, y_train = util.load_dataset(train_path, add_intercept=False) + x_train, y_train = util.load_dataset(train_path, add_intercept=True) # *** START CODE HERE *** model = PoissonRegression(step_size=lr, eps=1e-5) model.fit(x_train, y_train) - x_eval, y_eval = util.load_dataset(eval_path, add_intercept=False) + x_eval, y_eval = util.load_dataset(eval_path, add_intercept=True) y_pred = model.predict(x_eval) np.savetxt(pred_path, y_pred) From 9a737aa631e0ee63b8a58c19e155091963b0dfac Mon Sep 17 00:00:00 2001 From: Lorenzo Rossi Date: Tue, 5 Nov 2024 18:13:15 +0000 Subject: [PATCH 3/4] Restored version specification in environment.yml --- problem-sets/environment.yml | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/problem-sets/environment.yml b/problem-sets/environment.yml index 7d75146..eed253b 100644 --- a/problem-sets/environment.yml +++ b/problem-sets/environment.yml @@ -2,8 +2,8 @@ name: cs229 channels: - defaults dependencies: - - matplotlib - - numpy - - pip - - python - - scipy + - matplotlib=2.2.2 + - numpy=1.15.0 + - pip=10.0.1 + - python=3.6.6 + - scipy=1.1.0 From 26936f59f27488f10520645f4d7259603b645480 Mon Sep 17 00:00:00 2001 From: Lorenzo Rossi Date: Fri, 20 Dec 2024 11:40:43 +0000 Subject: [PATCH 4/4] Fixed bug: the function train_perceptron used to load the ds5_train.csv set as test set, instead of the proper ds5_test.csv set. Corrected this by replacing line test_x, test_y = util.load_csv('data/ds5_train.csv') with line test_x, test_y = util.load_csv('data/ds5_test.csv') --- .../PS2-5 Kernelizing the Perceptron.ipynb | 172 +++++++++--------- problem-sets/PS2/src/p05_percept.py | 2 +- 2 files changed, 89 insertions(+), 85 deletions(-) diff --git a/problem-sets/PS2/PS2-5 Kernelizing the Perceptron.ipynb b/problem-sets/PS2/PS2-5 Kernelizing the Perceptron.ipynb index 16c4ad5..3a30846 100644 --- a/problem-sets/PS2/PS2-5 Kernelizing the Perceptron.ipynb +++ b/problem-sets/PS2/PS2-5 Kernelizing the Perceptron.ipynb @@ -14,30 +14,36 @@ }, { "cell_type": "markdown", - "source": [ - "#### (a)" - ], "metadata": { "collapsed": false, "pycharm": { "name": "#%% md\n" } - } + }, + "source": [ + "#### (a)" + ] }, { "cell_type": "markdown", - "source": [ - "##### i." - ], "metadata": { "collapsed": false, "pycharm": { "name": "#%% md\n" } - } + }, + "source": [ + "##### i." + ] }, { "cell_type": "markdown", + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%% md\n" + } + }, "source": [ "Recall the update rule\n", "\n", @@ -49,28 +55,28 @@ "\\theta^{(i)} & = \\sum_{j = 1}^{i} \\beta_j \\phi (x^{(j)}) \\\\\n", "\\theta^{(0)} & = \\vec{0}\n", "\\end{align*}" - ], + ] + }, + { + "cell_type": "markdown", "metadata": { "collapsed": false, "pycharm": { "name": "#%% md\n" } - } + }, + "source": [ + "##### ii." + ] }, { "cell_type": "markdown", - "source": [ - "##### ii." - ], "metadata": { "collapsed": false, "pycharm": { "name": "#%% md\n" } - } - }, - { - "cell_type": "markdown", + }, "source": [ "\\begin{align*}\n", "h_{\\theta^{(i)}} (\\phi (x^{(i + 1)})) & = \\mathrm{sign} ((\\theta^{(i)})^T \\phi (x^{(i + 1)})) \\\\\n", @@ -78,28 +84,28 @@ " & = \\mathrm{sign} \\big( \\sum_{j = 1}^{i} \\beta_j \\langle \\phi (x^{(j)}) , \\phi (x^{(i + 1)}) \\rangle \\big) \\\\\n", " & = \\mathrm{sign} \\big( \\sum_{j = 1}^{i} \\beta_j K(x^{(j)}, x^{(i + 1)}) \\big)\n", "\\end{align*}" - ], + ] + }, + { + "cell_type": "markdown", "metadata": { "collapsed": false, "pycharm": { "name": "#%% md\n" } - } + }, + "source": [ + "##### iii." + ] }, { "cell_type": "markdown", - "source": [ - "##### iii." - ], "metadata": { "collapsed": false, "pycharm": { "name": "#%% md\n" } - } - }, - { - "cell_type": "markdown", + }, "source": [ "\\begin{align*}\n", "\\theta^{(i + 1)} : & = \\theta^{(i)} + \\alpha \\big( y^{(i + 1)} - h_{\\theta^{(i)}} (\\phi (x^{(i + 1)})) \\big) \\phi (x^{(i + 1)}) \\\\\n", @@ -110,29 +116,29 @@ "Therefore, the new update rule is:\n", "\n", "$$\\beta_{i + 1} := \\alpha ( y^{(i + 1)} - \\mathrm{sign} \\big( \\sum_{j = 1}^{i} \\beta_j K(x^{(j)}, x^{(i + 1)}) \\big) )$$" - ], - "metadata": { - "collapsed": false, - "pycharm": { - "name": "#%% md\n" - } - } + ] }, { "cell_type": "markdown", - "source": [ - "#### (b)" - ], "metadata": { "collapsed": false, "pycharm": { "name": "#%% md\n" } - } + }, + "source": [ + "#### (b)" + ] }, { "cell_type": "code", "execution_count": 1, + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%% code\n" + } + }, "outputs": [], "source": [ "def initial_state():\n", @@ -142,17 +148,17 @@ " An array of tuples. Each tuple comprises a training example and the corresponding `beta`.\n", " \"\"\"\n", " return []" - ], + ] + }, + { + "cell_type": "code", + "execution_count": 2, "metadata": { "collapsed": false, "pycharm": { "name": "#%% code\n" } - } - }, - { - "cell_type": "code", - "execution_count": 2, + }, "outputs": [], "source": [ "def predict(state, kernel, x_i):\n", @@ -167,17 +173,17 @@ " Returns the prediction (i.e 0 or 1)\n", " \"\"\"\n", " return sign(sum(beta * kernel(x, x_i) for beta, x in state))" - ], + ] + }, + { + "cell_type": "code", + "execution_count": 3, "metadata": { "collapsed": false, "pycharm": { "name": "#%% code\n" } - } - }, - { - "cell_type": "code", - "execution_count": 3, + }, "outputs": [], "source": [ "def update_state(state, kernel, learning_rate, x_i, y_i):\n", @@ -192,46 +198,48 @@ " \"\"\"\n", " beta_i = learning_rate * (y_i - sign(sum(beta * kernel(x, x_i) for beta, x in state)))\n", " state.append((beta_i, x_i))" - ], - "metadata": { - "collapsed": false, - "pycharm": { - "name": "#%% code\n" - } - } + ] }, { "cell_type": "markdown", - "source": [ - "#### (c)" - ], "metadata": { "collapsed": false, "pycharm": { "name": "#%% md\n" } - } + }, + "source": [ + "#### (c)" + ] }, { "cell_type": "markdown", - "source": [ - "Copy over the remaining code from `p05_percept.py` (no change on logic):" - ], "metadata": { "collapsed": false, "pycharm": { "name": "#%% md\n" } - } + }, + "source": [ + "Copy over the remaining code from `p05_percept.py` (no change on logic):" + ] }, { "cell_type": "code", "execution_count": 4, + "metadata": { + "collapsed": false, + "pycharm": { + "name": "#%% code\n" + } + }, "outputs": [ { "data": { - "text/plain": "
", - "image/png": "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\n" + "image/png": "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", + "text/plain": [ + "
" + ] }, "metadata": { "needs_background": "light" @@ -240,8 +248,10 @@ }, { "data": { - "text/plain": "
", - "image/png": "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\n" + "image/png": "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", + "text/plain": [ + "
" + ] }, "metadata": { "needs_background": "light" @@ -309,7 +319,7 @@ " for x_i, y_i in zip(train_x, train_y):\n", " update_state(state, kernel, learning_rate, x_i, y_i)\n", "\n", - " test_x, test_y = util.load_csv('data/ds5_train.csv')\n", + " test_x, test_y = util.load_csv('data/ds5_test.csv')\n", "\n", " plt.figure(figsize=(12, 8))\n", " util.plot_contour(lambda a: predict(state, kernel, a))\n", @@ -318,33 +328,27 @@ "\n", "train_perceptron('dot', dot_kernel, 0.5)\n", "train_perceptron('rbf', rbf_kernel, 0.5)" - ], - "metadata": { - "collapsed": false, - "pycharm": { - "name": "#%% code\n" - } - } + ] }, { "cell_type": "markdown", - "source": [ - "For the product kernel, there is no feature mapping for the input attributes. In other words, the model is still linear\n", - "after applying the product kernel. As a consequence, it performs badly on the data that are not linearly separable." - ], "metadata": { "collapsed": false, "pycharm": { "name": "#%% md\n" } - } + }, + "source": [ + "For the product kernel, there is no feature mapping for the input attributes. In other words, the model is still linear\n", + "after applying the product kernel. As a consequence, it performs badly on the data that are not linearly separable." + ] } ], "metadata": { "kernelspec": { - "name": "cs229", + "display_name": "cs229", "language": "python", - "display_name": "cs229" + "name": "cs229" }, "language_info": { "codemirror_mode": { @@ -361,4 +365,4 @@ }, "nbformat": 4, "nbformat_minor": 0 -} \ No newline at end of file +} diff --git a/problem-sets/PS2/src/p05_percept.py b/problem-sets/PS2/src/p05_percept.py index b69405b..b8e7e33 100644 --- a/problem-sets/PS2/src/p05_percept.py +++ b/problem-sets/PS2/src/p05_percept.py @@ -101,7 +101,7 @@ def train_perceptron(kernel_name, kernel, learning_rate): for x_i, y_i in zip(train_x, train_y): update_state(state, kernel, learning_rate, x_i, y_i) - test_x, test_y = util.load_csv('../data/ds5_train.csv') + test_x, test_y = util.load_csv('../data/ds5_test.csv') plt.figure(figsize=(12, 8)) util.plot_contour(lambda a: predict(state, kernel, a))