{
    "info": {
        "author": "Ruochi Zhang",
        "author_email": "zrc720@gmail.com",
        "bugtrack_url": null,
        "classifiers": [
            "Intended Audience :: Developers",
            "License :: OSI Approved :: MIT License",
            "Natural Language :: English",
            "Programming Language :: Python :: 3.4",
            "Programming Language :: Python :: 3.5",
            "Programming Language :: Python :: 3.6",
            "Programming Language :: Python :: 3.7"
        ],
        "description": "# SKNet\n\n## Introduction\nSKNet is a new type of neural network that is simple in structure but complex in neuron. Each of its neuron is a traditional estimator such as SVM, RF, etc.  \n\n## Fetaures \nWe think that such a network has many applicable scenarios.  \n- We don't have enough samples to train neural networks. \n- We hope to improve the accuracy of the model by means of emsemble. \n- We hope to learn some new features. \n- We want to save a lot of parameter adjustment time while getting a stable and good model.\n\n\n## Installation\n\n```python3\npip install sknet\n```\n\n\n## Example\n\n### Computation Graph\n\n![](./computation_graph.png)\n\n### Code\n\n```python\nfrom sknet.sequential import Layer,Sequential,SKNeuron\n\nfrom sklearn.ensemble import RandomForestRegressor\nfrom sklearn.ensemble import AdaBoostRegressor\nfrom sklearn.ensemble import ExtraTreesRegressor\nfrom sklearn.svm import LinearSVR\nfrom sklearn.linear_model import LogisticRegression\nfrom sklearn.ensemble import GradientBoostingRegressor\nfrom sklearn.neighbors import KNeighborsRegressor\n\n\nfrom sklearn.datasets import load_breast_cancer\nfrom sklearn.model_selection import train_test_split\n\n\ndata = load_breast_cancer()\nfeatures = data.data\ntarget = data.target\n\nX_train, X_test, y_train, y_test = train_test_split(features, target, test_size=0.2, random_state=42)\n\n\n\nlayer1 = Layer([\n    SKNeuron(RandomForestRegressor,params = {\"random_state\": 0}),\n    SKNeuron(GradientBoostingRegressor,params = {\"random_state\": 0}),\n    SKNeuron(AdaBoostRegressor,params = {\"random_state\": 0}),\n    SKNeuron(KNeighborsRegressor),\n    SKNeuron(ExtraTreesRegressor,params = {\"random_state\": 0}),\n])\n\nlayer2 = Layer([\n    SKNeuron(AdaBoostRegressor,params = {\"random_state\": 0}),\n    SKNeuron(LinearSVR,params = {\"random_state\": 0}),\n])\n\nlayer3 = Layer([\n    SKNeuron(LogisticRegression,params = {\"random_state\": 0}),\n])\n\n\nmodel = Sequential([layer1,layer2,layer3],n_splits = 5)\ny_pred = model.fit_predict(X_train,y_train, X_test)\nprint(model.score(y_test,y_pred))\n\n\n# acc = 0.9736842105263158\n```\n\n## Todo\n- Two or three level stacking\n- multi-processing\n- features proxy",
        "description_content_type": "text/markdown",
        "docs_url": null,
        "download_url": "",
        "downloads": {
            "last_day": -1,
            "last_month": -1,
            "last_week": -1
        },
        "home_page": "https://github.com/zhangruochi/SKNet",
        "keywords": "stack,sklearn",
        "license": "MIT License",
        "maintainer": "",
        "maintainer_email": "",
        "name": "SKNet",
        "package_url": "https://pypi.org/project/SKNet/",
        "platform": "any",
        "project_url": "https://pypi.org/project/SKNet/",
        "project_urls": {
            "Homepage": "https://github.com/zhangruochi/SKNet"
        },
        "release_url": "https://pypi.org/project/SKNet/0.0.1/",
        "requires_dist": null,
        "requires_python": "",
        "summary": "a library used for stacking based on scikit-learn",
        "version": "0.0.1"
    },
    "last_serial": 5703645,
    "releases": {
        "0.0.1": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "07921edaff56bcc6bf221e70196adb93",
                    "sha256": "ffcb669c4a20230b3ccef30ca837915ae0b14bbe13129ac97ec66cb4cca3fc11"
                },
                "downloads": -1,
                "filename": "SKNet-0.0.1.tar.gz",
                "has_sig": false,
                "md5_digest": "07921edaff56bcc6bf221e70196adb93",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 3240,
                "upload_time": "2019-08-20T14:28:42",
                "url": "https://files.pythonhosted.org/packages/7e/1a/b533ba48054794b41898f4cccd18387586a6b4c568975e696c3ae5692e71/SKNet-0.0.1.tar.gz"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "07921edaff56bcc6bf221e70196adb93",
                "sha256": "ffcb669c4a20230b3ccef30ca837915ae0b14bbe13129ac97ec66cb4cca3fc11"
            },
            "downloads": -1,
            "filename": "SKNet-0.0.1.tar.gz",
            "has_sig": false,
            "md5_digest": "07921edaff56bcc6bf221e70196adb93",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": null,
            "size": 3240,
            "upload_time": "2019-08-20T14:28:42",
            "url": "https://files.pythonhosted.org/packages/7e/1a/b533ba48054794b41898f4cccd18387586a6b4c568975e696c3ae5692e71/SKNet-0.0.1.tar.gz"
        }
    ]
}